This is interesting. I measured a higher clock deviation at 96kHz and 192kHz, than at 44.1kHz. I don't recall 48kHz, I'll check later. This is not something that we can hear, but that can disrupt a digital communication between two devices, if one does not properly slave to the other.
Also, the drift increases with temperature, up to 2 hours, after which it is stable. So, a communication can be perfectly working for the first minutes, and then will get worse as time goes on and temperature goes up. This particularly true for the digital SPDIF input, and as I’m not familiar with USB type transmission protocol, I don’t know in that case.
I'd be you guys, I'd try not to force resampling from the computer, if you can do that. Say if you listen to an original 44.1kHz sampling rate file, keep it this way with the player of your computer to the Topping. I'm not sure it will help, but if yes, it'd be nice to know from your feedback.
I did not experience issues with USB so I can't replicate/investigate. I had it with optical input, and I guess I found the root cause for this one.
Anyways, I now have a ton of measurements and that's enough to write a comprehensive review. There will be more to test and discuss, but there are lots of functionalities to review and assess, especially when some show flaws. So I guess it's better I release the review with some tangible data and we build on it, to later complement with ideas from the community. I’ll release it tomorrow.