I'm not a sound science guy so don't tear me apart.
The truth probably lies somewhere in between. It's like when objectivists try to say all DACs some the same (or even all properly implemented DACs sound the same). If someone can sit down with two properly implemented DACs in the same system and get 10/10 on a blind ABX test, some objectivists would still say there's no difference because there's numbers to prove it. At that point, no matter what the numbers say, it seems that at least the TESTEE can hear the difference.
The same can be applied to EQing, I'm personally don't EQ because I'm lazy, I don't know without messing up the sound, I think it colors the sound, blah blah all that BS excuse. Is that the same as saying I would never enjoy an equalized sound? No. Put me in a sterile system (which I have now) and a few hip-hop songs and I'd be begging for some EQing...
But there's a limit to it, a common-sensicle (not a real word, but a cool word non the less) limit. You can put up all the numbers you want but shove a SR60i driver into a HD800 shell and compare it with a real HD800. EQ your behind off and I'd bet $1000 that half a dead and blindfolded rabbit can tell the difference.
Another thing is, lack of bass/mids/highs are not the only problems that exist in the headphone would, there sound stage to consider, comfort, reasonance in closed cans, all kinds of things that can affect the sound of a headphone or speakers.
EQ does get you places, it'll save a lot of people a TON of money, but it is not the be all and end all of all your problems and even if someone can throw the numbers at you and yet you can HEAR the difference (not see, hear) - forget the damn numbers.
I don't think anyone's trying to suggest you can turn an SR60i into an HD800. Equalization won't improve a headphone's transient response or change the cup's design and resonance. It will change the volume of different frequencies. It's a method that should be employed once you get a headphone with everything but a nice sound signature.
Has someone actually been able to tell two "perfect" DACs apart in a blind test?