^I downloaded, carefully isolated a 5.435 second section, and then analyzed that "This guy is wrong" guy's first "demo", he plays 3 minutes into his video, a few days ago.
He either mistakenly (or deliberately?) botched the levels of the 4 test sequences I measured, by either average levels in Audacity or psychoacoustically-corrected level analysis such as ReplayGain,
so of course they "sound" different for that reason alone! Sheesh, this is the oldest trick in the book in audio!
My analysis shows the levels when compared correctly—rather than simplistically using just peak level analysis—are all over the map considering we can hear small differences of even just a half dB, or so, at least under the right conditions:
[Individual 5.435 sequences I isolated (and time-synchronized) available upon request.]
His test was also
sighted, not blind like it should've been, and some of his preamps even have his name on them
flashed on screen as you hear them so that can certainly sway people subliminally.