None whatsoever, actually, with either music (1938 Karajan/Berliner Philharmonic recording of the overture to "The Magic Flute"), or the "Ultimate Headphone Test" files; any difference made by removing the foam is evidently so subtle as to be completely lost on me. I don't feel like dismantling my headphones again to put the foam back right now, so I'll leave it out (stuck to the inside of a Zip-loc bag, since I don't have any proper backing paper handy to preserve the adhesive) until the next time I take the pads off for cleaning, and put the foam back in then. I doubt I'll notice any more difference in sound at that point than I do now.
> I wouldn't do it because I'm a ham-fisted klutz and I'd probably damage something.
On the other hand, the modification really is as simple as its adherents purport it to be, especially since the cord is detachable; the only even vaguely dicey moment I had was when I took out the left driver and found that its connection to the cord socket PCB is made by a wire that's both quite short and very fine, and that the foam would have to come out past the wires to both drivers. If I'd dropped the driver, I'd probably have needed to trim the wire and heat up my iron, so I'm glad I didn't do that. Other than that, it was a snap; anyone can do it in ten minutes who has screwdriver, tweezers, and reasonably steady hands.
Edit: Now I think about it, what I really should have done was take the foam out of one side, but leave it in the other, and see if I could pick out any difference that way; having both sides in the same state makes it essentially impossible to compare. Perhaps, when it next comes time to clean my pads, I'll give that a try.
Whether that difference is real, or some psychological boost you get from the "free upgrade to better headphones" is debatable.
I wouldn't do it because I'm a ham-fisted klutz and I'd probably damage something. Other people are not as clumsy as me!