c-attack

No free lunch on this and that 4 way splitter is just that...  4 unpowered connection ports.  It's real purpose is to allow you run multiple independent chains back to the system hub without having to run them through each other.  Since your radiators are in side by side pairs, I don't see an advantage.  
 
You will need 2 CUE Link Hubs for 32 fans.  I theory you could do this with three to avoid the 14 fan brightness clipping, but I think a 10% reduction for going 2 fans over (16 per hub) is worth it and these QX are really bright to begin with.  One hub for each 480 radiator pair.  8 to one side and 8 to the other.  Pump will need to go into one of these chains.  I suppose that's what the 4 way splitter is for.  You can run the XD5 back to the splitter on one end rather than being forced from a fan end point to the XD5 to the hub.  Very case layout specific as to whether it helps.  I can't tell if that GPU is block is CUE link of not.  

Contact Corsair Support.  Assuming these are bad bearings on the QX, they will send you new fans.  Ask for the Advanced RMA.  This is where they ship you new parts on a credit card hold.  You get the new gear, swap it out and send it back.  Credit Card hold released.  

I don’t think there’s any reason the worry. The outer sleeve is cosmetic and you sometimes see small sections that are not ‘skin tight’ on the underlying tube. That little wave closer to the radiator side may relax in time after the tubes heat up (mostly from case temp). However, if you feel genuine kinks or creases in the underlying tube beneath the sleeve, that is potentially problematic. It could restrict the flow and those should not pass a QC check. 

Go ahead and start a Support Ticket.  There's no way for any of us to be sure about the cause of the problem.  However, over the 9 months this product has been live, nearly all issues of this type have been user install errors or a bad cable.  You've tried a different hub, you do have some functionality with it, but it's not picking up all the devices. It's not going to be some type of unique software error that only affects you.  All I can recommend is looking at a lot of online videos to see how they have connected things.  

The things saved to the KB are still there and should automatically populate on the new PC once you connect the KB with CUE running.  However, those are HW files and they don't carry over into software profiles.  Those are the things you create in CUE.  You needed to export them from the main menu on your old PC.  If you still have it, you can still do it.  Data is saved even if the devices are not still connected.  These exported profiles can be saved to a flash drive and then re-imported into your new PC.

That is all normal and the CPUID CUE uses to fetch the data has trouble with the Asus Embedded Controller.  Other programs like HWinfo will give you a warning about it, but seem to handle the interaction better.  You get lots of duplicates and junk values at the min/max end of the range.  This has been an issue for many, many years.  
 
Similar discussion earlier this week.
 

Yes, the 4000X is ATX motherboard compatible.  
https://www.corsair.com/us/en/p/pc-cases/cc-9011204-ww/icue-4000x-rgb-tempered-glass-mid-tower-atx-case-black-cc-9011204-ww#tab-techspecs
 

The “H100” (in whichever form you have) is not a CUE device. AIO units that are part of the CUE control system have an “i” at the end of their number (H100i Elite Capellix). The H100 RGB was specifically designed to pass controls over to the motherboard.

Unfortunately I don’t know any quick fixes and it does suggest a physical hardware issue. What some users try is cleaning the scroll wheel contact point. You might be able find a video tutorial online (not from Corsair). I’ll look around later and see if I can find one. I’ve got the same issue with a much newer Dark Star mouse that suddenly started doing this, but none of the other Corsair mice I tested had the problem on this CUE version so it’s likely a hardware issue as well. If you are over 2 years then you likely are out of warranty but there are differences in some countries to comply with local laws. If you’re under 2 years and still in the warranty period, I’d go that route rather than disassemble unless you feel confident. 

The first video section is about the expected normal for a high speed AIO pump.  Unlike pumps from a few generations back, we've moved into narrower CPU channels for better cooling and it requires more flow pressure as a result.  The last several Corsair pumps all have a minimum speed just above 2000 rpm to account for that.  It's a high speed device, mounted to a motherboard, which in most cases has tendency to amplify the motor vibrations.  I am not sure I would classify any of the current AIOs across the market as "dead silent" in a quiet room at close range.  They are not physically capable of that.  You need to go to custom cooling with different pump design and mounting system to get that.  
 
The second part of the video is abnormal.  Most likely some back pressure from a large bubble somewhere, but it's possible there is a mechanical aspect.  With the pump at Extreme (max), lift up the front of the case and hold it for 10-20 seconds to see if you can release any air trapped in the pump area up and out of the block.  It's fine if it settles elsewhere, but around the pump or CPU channels is problematic for noise or temps.  If this does not resolve it or it continues to be persistent, contact Corsair for replacement or through the vendor.  Do not wait for it to 'get better' someday.  That's too far off and not worth the interim aggravation.  

Contact Corsair Support. This isn’t a user addressable problem and generally requires hardware replacement.  The pump fail warning may be caused by a bad tach sensor but it effectively locks you out of the device. 

You can try a different charging cable if you have one that fits, but you need to contact Corsair Support through the link at the bottom of the page or through the main website.  Nothing we can help you with if the battery is not able to hold or receive charge.  

If the coolant shutdown triggered, then the pump stopped and that should never happen -- even if the pump loses it's USB connection.  It might happen during a FW update if the process stalls and lingers for too long.  The pump does need to power cycle to complete the task and that is also what usually restores the communication rather than the FW re-save.  
 
Any time the pump stops you should take it seriously.  Start a Support Ticket with Corsair.  Even if it does not happen again, the ground work is laid out.  On that note, they are going to ask you to do the power cycle, re-connect the Commander Core to pump 22 pin power cable, probably clean install CUE, and do you have any other RGB control programming running, etc.  Try and get out in front of that in the ticket by stating you've already done the power cycle trick and so forth.  I don't see any evidence or reason to believe this is a software conflict with another program.  That usually presents a whole bunch of other issues like garbage data, impossible temps (-52C), etc.  Nevertheless, if you feel like doing a clean install while you wait for them you can.  Make sure you export any valuable profiles first and this is also a good idea if they are going to ship you a new unit.  You'll want to reimport the profiles a new AIO, otherwise it will think you are running 2 and not apply the old profile data.  
 
https://help.corsair.com/hc/en-us/articles/360025166712-iCUE-How-to-perform-a-clean-reinstallation-of-iCUE
 

It seems like your MB side wall and bottom fans are intake and then force exhausting all case waste heat out through the top panel where the radiator resides.  This is going to lead to higher than usual coolant temps, particularly in gaming when the GPU is spitting out 300W+.  You can try to alternate intake/exhaust arrangements: 1) Turn the MB side wall fans to exhaust to see if you can draw more GPU heat out through that path vs the top exhaust.  2) Move the radiator to the side wall as intake or exhaust (case/room position dependent).  Keep top as free air exhaust.  This should help keep more environment waste heat away from the radiator.
 
As for the immediate set-up, +1C to coolant temp = +1C to CPU temp.  The only disadvantage of 41C vs 39C coolant is it makes the CPU +2C warmer.  If your CPU temps are hitting your personal limit then you'll need to get strong on the fans, but otherwise set them to run at noise tolerable levels.  Int his case, I think all you really need to do is create a 50% PWM plateau in the 39-44C range.  This is where you'll be most of the time and this seems to be a tolerable speed for you, just below the annoyance level.  Leave the heavy fan blast at 45 and 50C.  If you hit 50C, you need to know.  That should be out of bounds for most everyone.
 
As to "typical temps", most user will idle 4-7C above room temp and see gaming temps +10C above that.  It's the first part where you are losing some ground.  Being able to keep the coolant baseline ~7C lower will let you further relax fans speeds or lower CPU temp 7C across the board.  

You might need to clarify your intent a little.  However, for custom water cooling any D5 or DDC pump can easily handle the flow resistance from two CPU blocks or a CPU and GPU block (if that is what you mean).  What you may want to account for the approximate amount of power you need to dissipate vs the number of radiators.  Most water cooling kits are not "bundle deals", so don't be afraid to choose what you need on a piece by piece basis.  No such thing as "guaranteed to fit all cases and installations".  That's why it's custom cooling.  

Inlet/Outlet to the front or back will not matter if the radiator is up top.  Any air will be pumped out of the CPU block.  You do not want to turn the inlet/outlet toward the bottom.  Aside from GPU collisions and temperature, that does create a pocket for air to become trapped in the CPU block.  

There is no mechanism to move settings from hardware modes to software modes or vice versa. Each mode has different device specific rules related the memory storage and execution, so freely transporting the more expansive software capabilities would create problems. The first time you set up, you’ll need to create your options in both modes, if you need the operations both when CUE is running and when it is not. After that, you should be able to copy hardware settings to other hardware profiles and standard software based operations to other software profiles using their separate libraries. 

Yes, if CPU fan is convenient for connection purposes, use that. The 3 pin connector from the AIO is actually just 1 wire — a tachometer. The AIO does not need this connection at all and its real purpose is to satisfy the standard motherboard “cpu boot protection” system. If you don’t connect something to cpu fan, the motherboard will throw up a warning and prevent you from booting. This is an old system from the standard air cooler days and prevents a novice user from booting up without the little blower fan turned on. For your purposes, it acts like a first line warning system if the pump does not talk back to the motherboard on power on. 
 
On Asus boards (and most others), AIO_Fan and W_Pump are just a chassis fan header preset to 100%. This is useful for an AIO that gets its power from the motherboard. However, most AIOs get their power from a SATA connection these days and thus the bios speed settings irrelevant and there is no control wire. Also, on most Asus boards AIO Pump can be unlocked turned back into a CHA_Fan header in the BIOS so you can use it for case fans. CPU fan and its copycat OPT always will be tied to cpu temp and be more reactive than any CHA fan header. Even with fan delays set to maximum, fans connected to cpu/opt will change dynamically in response to sudden cpu temp change. You don’t need that for radiator or case fan control. 
 
If AIO_Fan is more convenient for cable management, you can use it. The AIO has its own warning system for problems and will blink red in the event of pump failure, along with the unmistakable rush of fans up to maximum 30 seconds later. It also will indicate this in the software, so the BIOS level cpu protection is not required. However, you will need to disable it. Go to the Advanced BIOS  (F7) > Monitoring column. Find the CPU fan RPM line. Hit enter and choose “ignore”. That will turn off the boot protection feature when gpu fan is empty. 

There are a couple of ways you can try and address it.  
1) 3rd party adapter for the motherboard style ARGB connector.  This will allow it to properly connect but it does pass the control to the MB RGB program, which may or may not be useful.  That one fan will not change when CUE profiles and lighting do, so this really does separate it.  Really only viable if you don't change themes very often and use common effects like static colors, rainbows, etc.  
2) Corsair specific fan RGB splitter - This will "twin" two of your fans connected to the Commander Core and they will share a Y-splitter to one port.  This will keep all the RGB in the Corsair app and change with everything else.  However, if you do moving patterns like visor or any kind of wave effect, the two fans on the Y-splitter will display they effect at the same time.  The two fans light and act like one.  This means you need to choose the paired fans carefully for the least annoyance.  This is a non-issue for static or intra-fan effects like spiral rainbow, rainbow wave, etc. since all fans in the circuit do the same thing at the same time.  That makes this usually the best choice for most people.  This is an example of Corsair specific RGB splitter for a RGB fan port.  You can do the same thing on the PWM side with any common PWM 4+3 splitter.  

Device to CUE communication is along the USB 2.0 line from AIO to motherboard (internal).  See if Windows detects it as device.  Some motherboards are kind of finicky with their USB 2 connections.  Try a different USB2 port if you can.  There usually are two.  

Try running a repair install of the program.  Windows Settings> Apps> Corsair iCUE.  Click the drop down menu and select "modify".  Follow the prompts for a repair installation.  This does not erase profiles or settings but is good at cleaning up weird issues between OS and CUE.  There have been some rolling issues with the CPUID sub-program and that is how CUE gets that data. 

Obviously loss of fan control is annoying, but if the fans are maxed or off, that won’t cause the CPU to overheat. Only a pump shutdown will cause that on boot up. If the software can’t communicate with it or it’s not installed, the AIO runs from its onboard stored profile. There is no setting in CUE to turn off the pump.  All AIO power on that model comes from the motherboard header. Does that header show any RPM reading in the BIOS?
 
CUE 5 should work with a GTX and it’s hard to tell why things went wrong. However, if you have CUE 5 installed, Link won’t work and I seem to recall one post in the past year where someone jumping from old Link to CUE had issues caused by the dual installation. These programs are three generations apart and not meant to be run simultaneously. Make sure you remove one or the other before continuing. Pick the interface you like better.  If this is your only cue device, there won’t be too much difference in functionality. 

No, that’s what the shade does. It does not matter what effect(s) are underneath. The top layer static black shade will let varying amounts of light through depending on the opacity. That’s a 0-100 step brightness slider. 
 
You also can set the rgb values for the color shift and if your using lighting link with a random color pattern, use the shade — exactly as in the snapshot above. 

Prior versions don’t exist with the modular server side CUE 5. It’s also not likely this issue is version specific, but rather an unintentional misread on the new install. 
 
A couple of things you can try:
1) Re-run the rgb wizard for the Commander XT. That controller has independent rgb ports and turns off those it thinks are not in use. There may have been an error in the post install detection. 
 
2) Shut down the pc and flip the PSU off. Hold down the case power button for 10-15 seconds to drain any residual power. Then flip the PSU back on and start normally. This will force the Com XT to recheck all connections. 
 
3) Repair the software - Windows Settings> Apps> Corsair iCUE. At the end of the line choose modify from the drop down menu. Follow the prompts for a repair install. This will not erase profiles or settings. It may correct this issue if it’s a corrupted lighting config file. 

Ok, that’s not necessarily bad. It means the software has never seen the hub, so we are looking for a motherboard side reason or a physical problem with cable rather than CUE misbehavior.  Could be the hub too, but your swap puts that at extremely low odds. However, if you’re not getting anywhere grab a different usb cable from one of the other hubs.  Also, if this is a brand new build, make sure you’ve installed the chipset drivers from the board maker (or Intel/AMD). Those handle usb traffic.