I've just taken a look at that link. It cites the j-n-d as being the sort of tool used to develop processes such as lossy compression such that the losses remain below the j-n-d threshold.
Without wishing to light any fires on here, I am assuming that there is broad agreement among us that lossy compression (eg MP3) is clearly audible.
Does that suggest, perhaps, that there are different levels for the j-n-d and that, as audiophiles, we may be in a minority which perceives those differences, where the majority either don't, or don't consider them important. After all, we go to some trouble and expense to get a higher than average standard of replay quality for our music, and the great majority of people don't, and don't seem to care. That might suggest we perceive something as important which many do not.
I'm not suggesting any sort of golden-ears, merely that there is invariably a continuum in most aspects of human behaviour, a sort of bell-curve of typical and atypical behaviour and response. Perhaps audiophiles are skewed towards one end of that bell-curve as far as music perception is concerned.