+1 @ghdprentice , I just moved from a BMC Ultra DAC to a Meitner MA3 and as an experiment, I removed my Etherregen, AD clock and the AD LPS powering them. Turns out, to me, sq improved quite a bit -- smoother, more musical and more detailed, with a marked decrease in the occurance of excessively high or shrill passages in certain recordings and alto sax runs. This may have had something to do with my now runing ethernet straight from the LPS powered modem and Bonn8 switch to the Zenith MK3 and ethernet, not USB as with the Ultra DAC, from the MK3 into the MA3. I use high quality ethernet cable in both instances (Supra CAT 8 and Veritas Mirus).
Everything matters
I have now been tinkering for 10 years with my streaming set up and the principal lesson learned is that everything matters. Think about it: from the wall outlet to the router to one or more switches to the streamer and dac, the latter three dependendent on clocking and all of them depending on cables and power supplies), it stands to reason that e.g. changing a USB cable is very likely to compensate for a weakness in the chain, that if addressed, would result in an overcompensation by that cable. Therefore I look at individual component or cable reciews with a very jaundiced eye. It is the performance in the context of the individual chain, that matters. As a result I have had numerous ethernet or USB tweaks, be the filters, isolators or processors, cables and LPSs in and out of my chain. For what it‘s worth, the biggest impacts ranked on SQ I experienced were from inserting 1. Etherregen, 2. LHY Audio UIP, 3. 10m clocking of Dac, UIP and Etherregen and 4. LPSs on router, switch and Etherregen.
My main point though remains that praising individual components irrespective of the digital chain is borderline meaningless