Compared to results from the same system on the open test platform, the GPU results were nearly identical, with the range being 76 – 77 C in ambient temps of 21.5 – 22.0 C. However these temps are targeted by the GPU, so average clock speeds will be a better determination of performance here (more on this in a moment).
As to CPU temps, I was going to create an open testbed vs. 4000D chart until I realized that all testbed GPU temps were recorded using 100% CPU fans to prevent thermal bottlenecks with our processor. Still, load temps topping out at around 75 C with a Ryzen 9 3900X are fine, though I probably would have had lower numbers with the radiator positioned on the front intake – drawing in cool air – rather than the top (which would also have likely affected GPU temps).
(For the record, the same CPU/cooler on the open testbed with 100% CPU fans was only 41 C under this simulated gaming load. If you can live with the noise, the H100i PRO RGB is extremely effective with high fan speeds.)
To see if GPU clocks were affected to any significant degree inside the 4000D vs. open testbed, I created this chart using the GPU clocks from GPU-Z from the 10x consecutive Metro Exodus benchmark runs, both in and out of the case.