I'm not sure I saw John S. describe it as affecting the audio clocks but rather he wrote that the increased processing at the USB PHY caused power and current spikes affecting the D/A at the DAC chip. Hence he worked on the power-related noise profile of that USB PHY section in the Regen (among other things).
I appreciate you sharing your additional info based on your own experiments, very interesting.
At least you understood correctly where the theory of operation was focused, contrary to some who claimed recently that it was based on asynch USB re-sending data (LOL at that!).
Thanks, I do understand the theory & appreciate where Swenson is coming from & it roughly concurs with my own theories as a result of my experience & experiments.