Thank you, Mario. I am grateful for your experiments and observations.
The interesting thing from the m2p perspective is that the apparent hardware problems are experienced by m2p software but not volumio nor moOde. And that the problems with power have such a narrowly observed misbehavior. I would bet that it is possible to code the m2p operating system so that this does not happen. I am not saying that would be necessary nor straightforward, though.
The interesting thing from a Rpi perspective is that the cable makes a difference in your experience. In my configuration, I have a USB-C charger rated at 3 A, which is what the Rpi 4 B is said to require when allowing for extra current required by USB connections. My USB thumb drive does not have specifications on how much current it consumes. The HifiBerry Digi+ Pro that I am using is no longer produced, but the new Digi Pro is said to use a negligible amount of current, so I doubt that is part of the problem. I am suspicious that Rpi 4 B is not configured optimally for independence from cable parameters, where it could be. But this is only speculation.
From my perspective: I have ordered a Raspberry Pi 4 B power supply, 3.5 A, for further experimentation. The experiments are pretty time-consuming, and even if I show that using a different cable fixes the problem, I doubt it has a comfortable margin. So I want to use a more complete fix of a new power supply with integrated cable and more power than is needed.
I will let you know of the results. Thank you so much for replicating the problem.