I read it a while back, when Tyll linked it on IF. One of the grading criteria I disagree with is the thing about treble air. While air is very important, the level he prescribes is IMO too high, resulting in tizziness, grainy noise on strings, and a faintly artificial overall sound. I know this all too well since my SRH440, which incidentally would get a big fat FAIL on this metric based on the grading scale in that paper, has too much upper treble energy (especially in a peak centered around 15.5 kHz). It's amazing how much more natural they sound with a bit of precision EQ.
As sound travels through space, treble is attenuated. The very highest frequencies are lost first. Taking this as the model, there should actually be quite a bit of rolloff in the upper treble to produce a natural sound on headphones, since the transducers are right next to the eardrum and thus the sound doesn't travel far enough for natural treble rolloff to occur. The guy who wrote this paper might have come up with his air criterion based on at least one of the following two conditions:
1) An assumption that a perfectly flat measurement on Tyll's graph is an ideal result, and/or
2) He might suffer from high frequency hearing loss (either naturally from age or as a result of illness or hearing abuse)
I don't suggest #2 to be crass. Since the highest frequencies are usually the first to go, he might find a higher than average level of treble air to sound correct, as it compensates for any natural lacking in sensitivity he might have in this range. I apparently have an above average level of hearing range retention for my age, according to a test I took on HF a while ago (I can still hear up to 18 kHz), I certainly don't need any extra air, and I feel like the author's "ideal" would sound grossly out of balance to me.
I think these kinds of "ideal finding" discussions are at least as revealing of the perspective of the author as they are of the subject matter. To that end, I suppose we cannot exclude a third possibility, and that's the matter of personal preference. Some people really enjoy air, since it can add a sense of presence and space around sonic elements. It's one of the secrets to the DT880's magic, except that I don't feel the Beyers are excessive in their upper treble level.