One of my favorite set of pictures to show EQ doubters was that of the waveform of a test signal containing sharp transients. The first picture of the waveform, as recorded off a pair of headphones, showed that the phones were unable to keep up with the quick transients, getting them all blurred. For the second picture, proper EQ was applied, and now the same phones were able to reproduce the transients with absolute precision. Thus, whether the driver had gotten better or not, the end result was that the phones' ability to resolve detail had increased dramatically with EQ.
I hoped someone would've pointed out some error of logic in that set of pictures or the method behind it, but no one did. I'm not sure if people in general had even registered what they'd seen. More often than not, a person would maintain that EQ can't make headphones more resolving, see the pictures, then simply restate verbatim that EQ can't make headphones more resolving.
Not related to EQ, I was once working on a mod for some planar magnetics where the idea was to cover the driver with a thin sheet of plastic - like what's done with electrostats but never with planars - to protect the driver from dust. There was a guy in the thread where I was posting about this who speculated that plastic right in front of the driver would ruin the sound. I did the mod, listened to it, and found the sound unaffected. The guy maintained that the sound would be ruined. I took measurements - before and after - that showed the sound virtually unaffected by the mod. Having seen the measurements, the guy repeated that the mod wouldn't work because it would ruin the sound. I don't think he ever tried it.
In the same way, the beliefs of the guy with the HD 800 will always override your views that the HD 540 is a better headphone, no matter how many comparisons you do.