Of course soundstage is not exclusive to the brain, else every headphone would have the same soundstage given the same brain.
Localization is greatly influenced by frequency response differences between channels. We can measure changes in frequency response. There's no need to model a human brain to reconstruct how it will localize different instruments.
But since people make claims about the FR changing, this is an irrelevant point anyway.
The problem is that brains do not only incorporate physical sound waves (which we can easily measure) into what we hear, but also experiences, expectations etc. that bias our perception.
That's why you can trick people into hearing audible changes by fiddling with knobs on an equalizer that is disabled.
Edited by xnor - 1/9/14 at 4:43pm