I love when the cable threads meander into the realm of law and epistemology! (My grad school focus.)
The statement "I can hear a difference, therefore a difference exists" is not a sound argument, only a valid one. There are several reasons why the statement "I can hear a difference" may be true: you may perceive a difference through expectancy bias, for example. But you would need to prove that "a difference exists" independent of the statement "I can hear a difference" in order for the argument to be sound ("a factual argument").
I'm not claiming that "so-and-so can not hear a difference." Maybe so-and-so can. But for you to come to a sound conclusion that "a difference exists" requires a bit more legwork than merely asserting the truthfulness of your premise.
The test-that-shall-not-be-named is the only accepted way that I know of, outside of audiophilia, to determine the truthfulness of the conclusion "a difference exists."
The time honored question "If a tree falls in the woods where there is no one to hear it, does it make a sound?" plays upon the same ambiguity of meaning/reference. If you insist that only the second meaning of "sound", a vibration of the ambient medium between 20Hz and 20kHz, is correct, then you need another term for what is heard, since we all know that there are other factors making the vibration and what is heard often not tightly correlated. However, I contend that as audiophiles it makes more sense to use "sound" for what is heard and to call the vibrations the "source" or "signal" produced by the system.
While these discussions are usually pointless because of polarized positions and ulterior motives, they are always meaningless without keeping these distinctions and terminologies clear.