Indeed, this is why i suggest people do their own tests if at all possible.About 10 years ago, when jitter was an undisputed, real issue in digital audio, I bought a hardware upgrade for my Squeezebox Touch, which included new and better spec'd clocks and whatnot, costing about $500. I sent my Touch away for this procedure, but kept my second, unmodified Touch. When I got the modified Touch back, I played various test tracks, and immediately heard a clear and obvious improvement, particularly in the treble. This should be no surprise, given the better clocks etc. and the known audible difference jitter can introduce. However, just for fun, I plugged my unmodified Touch in next to the modified Touch and shifted between them to the same DAC (Audio-gd Ref 7.1), making it easy to hear any difference. In short, there was no audible difference, and this test was done with electrostatic speakers (see my profile for details), that should be resolving such a difference clearly
The R26 thread has page after page of people extolling the night and day improvements of the LAN connection over the USB and as for toslink you must be mad. When I did my own tests the toslink was indistinguishable from the LAN (neither myself nor my wife being able to detect difference what so ever). That doesn’t mean in some people’s setups their might be: Network setup and layout, quality of mains, emf environment etc etc all vary and may affect end results.
Expectation bias is difficult to overcome, I am not keen on tests with rapid switching between two sources as they often won’t tell you which you prefer, but they should tell you if there is any difference (not the same thing as which is preferable). I could easily distinguish differences between x26pro and R26, rapid switching showing more bass weight of x26 for example, but only more extended tests told be I find the R26 a more enjoyable listen.