that's intriguing for sure. I did a little test last year after some fellow got me curious about this. and I did get tiny changes here and there in the measurements, but keep in mind that I used one magnet I happened to have around. and it remained on top of the back plastic design, unlike you who murdered a piece of that to get closer to the driver. so we can probably expect your experiment to supersize everything I had.
here's what I got(FR up and impedance below). I have absolutely no idea if this is what one should expect from sticking magnets around so take it for what it is, me fooling around and having no idea what I was doing ^_^.
the little reduction in the bass seems kind of consistent with the impedance change, but for this I removed the pads(why the low end rolls off that much and why the overall FR doesn't look like much of anything we know). I really did my very best not to move the headphone between the takes, and the FR is close enough to assume that I did an OK job of not moving much. so the little change found at 4kHz is also probably due to the magnet somehow(but maybe not, I don't even have a vague idea to explain that). otherwise the response didn't seem to care much.
as for distortions, well I'm using crappy stuff so I always measure way more than I should and my measurements aren't reliable. all I can tell is that THD showed no significant change in my case, nothing big that pops out anywhere. the variations weren't bigger than just taking the same measurement twice, so if something changed in term of THD, my crappy graphs were inconclusive.
now for the curious kid inside me, I would really love to see how the magnetic field is affected. because ideally we'd like something fairly consistent through the course of the coil. and IDK if a bunch of little magnets placed in circle are going to do that? again I don't have a clue. to quote a famous philosopher: "fracking magnets how do they work?"
but it sure is something interesting, I'd love to have a headphone/speaker designer to talk on this.