The rads are 2 x 360/45 with 6 120mm cheapo fans... 16mm hard tubes and using the Lian li manifold (hooked up to the cryogenic plant and superconductors... temps went deep into the red last October) lol.
In normal use, it hasn't been switched off for 5 days now, running 5 searches on the NAS, security camera feeds and other stuff, (13 apps open and 10 windows) on 2 ultra wide monitors running 3440 x 1440... (40% peak GPU load) CPU 32c, GPU 29c. Coolant 26c, (16c ambient)
There are three additional temp probes fitted in the loop at the reservoir / manifold. Output from GPU (the same temp as the input to CPU). Output from CPU. Then through the 2 rads, via manifold back to reservoir. Fans and Pump are all at 100% but i can still hear the clock on the wall above it ticking, so the noise levels are perfectly acceptable given the case is only about 1 metre away.
The worst temp I'm seeing is 60c on the HDD in the rear hot swap cage.. I'll be adding another fan in that area to bring that temperature down a bit. It's not a brilliant design. The more I look at the Dynamic XL the less impressed I am with it. I will not be buying another one.
Version 2 of the loop is already in CAD, I'll break out the CNC and try some acrylic carving
... I don't like that the coolant has to go through the GPU to get to the CPU, I'd rather they were both fed straight from the reservoir with the coolest fluid in the loop. I have a design that will be able to do that. See the bit above with coolant temps going into the CPU being 3c higher than they need to be, (so the GPU, at the time of observation, is getting coolant 3c higher than it could be getting straight from the reservoir.)
Overall, all systems nominal.