It’s an error commonly made in evaluating hi-fi–system performance: the failure to listen differentially. Differential as in compared to something else. “Something else” could be a different recording on the same system or (especially this) the same recording on a different system. The question is, what are you comparing it to? The point is: Do you really know what that recording sounds like?
When it comes to evaluating equipment, audio engineers—especially those who specialize in naturalistic recordings—have an advantage. Someone like John Atkinson—I wrote “someone like,” though really he’s a category of one—knows better than anyone what his recordings sound what’s on the recording. Even so, even for him, all the evidence of what’s on there is indirect, via his monitoring equipment. The very notion of a recorded sound, independent of a reproduction system, is fraught.
