Could other characteristics of the sound count for nothing... eg. amplitude of peaks (as opposed to average volume)? I don't know the answer to this question, I am just curious since I could imagine a characteristic like that possibly being affected more by proximity/distance (peaks perhaps being damped by distance more than average volume/sound pressure?).Proximity in itself means nothing. The volume is the only thing that counts, which decreases with distance. This is basic physics and subjective arguments and beliefs are worthless.
I'd want to see some evidence before making any categorical statements, but something about listening to music on headphones does make me a little uneasy. I mean the experience of it, rather than any preconception. I use them sometimes for attempts at neuromodulation or frequency finding, but have just felt a bit uneasy when I've put music on them. I know that is subjective, but I don't want to reject the subjective completely since science can change as new factors are taken into account. Just like in audiophilia more generally, I suspect that subjective impressions can tell us things that 'objective' measurements do not. But that is another argument for another place. Or not.
Generally speaking I agree that volume of sound at the point of hitting the eardrum is by far the most significant factor, however the sound is generated.