Tuning is arbitrary. I've listened to some comparisons between A=432 vs 440 and I think the 440 actually sounds better. But then again, like this article even states, it depends so much on the piece. If you're listening to stuff from the 1700s/1800s, it would make sense to try to use the tuning they used at the time, if only to try to get the sense of emotion intended by the originally composers/arrangers/musicians (even then it's a best guess).
But yes, it is pretty arbitrary. It's the tuning of a note depended on the number of cycles per some arbitrary time unit we defined.
If you're listening to a song today, with all our sound precision technology, you're hearing it the way the original author/musicians wanted it to sound (depending on your equalizers, speakers and various other audio setup).
I'm curious how you can tell. A double-blind test spaced a few minutes apart (enough for you to forget the frequency so you couldn't tell if what you were hearing was tuned higher or lower than what you heard on the previous trial) would be interesting.
I play a lot at A=440, but I also play a lot at A=415 - mostly on the viola da gamba. A=415 is roughly one semitone lower, and it changes the way the instruments resonate. My viol happens to sound good at either pitch, but as an ensemble we find that when we're playing at A=440 the whole sound doesn't blend in quite the same satisfying way. This is due to the construction of the instruments themselves, but also just how those slightly lower pitches sound.
Oh and recorders built to A=415 just sound great. I think that slightly bigger body gives them the chance to produce a richer resonance, but then again you also don't find cheap plastic recorders (or cheap wooden ones) built at A=415 because only "serious" players go looking for those instruments.
One nitpick on the article though - historical tunings don't range from A=415, my recorder teacher has two flutes at what is sometimes called "French baroque pitch" - that's A=392!
The article had a pretty good explanation: that strings sound different at different tensions. Musical instruments tuned lower would sound darker, mellower, and "looser" than those tuned higher. The instruments would probably feel a little different to the musicians, who would probably play differently, especially if they'd just heard somebody expound on the mystical properties of a certain tuning. I'm not at all surprised that it would sound different.
On an FFT graph, you'll have the same fundamental and a very similar next few partials, but eventually you'll see (and hear) the results of the different tensions. Those differences represent what people like and don't like about different tunings -- nothing to do with the slightly different pitch between 440/442/438/432/etc!
Personal example: As a musician, A=432 doesn't sound that different to me, not enough to make that much of a difference in tone. However I do like A=425, an arbitrary "in between" tuning on the flat side of the middle between A and Ab in A=440 tuning. It adds an entire new mood set for each key that is rather different than A=440, distinct enough to me where the moods don't overlap (whereas with A=432 it does).
There is no science that can explain this sort of thing that I know of, and I fully expect that others would have different interpretations.