I don't
think Sean has specifically advocated having all headphones have the exact same response - certainly, Floyd hasn't advocated having all speakers have the exact same directivity. That level of...specificity in prescription is something I don't see in their work.
If I were forced at gunpoint to stylize the solution to the circle of confusion per what I understand Sean and Floyd to believe, it would be that all speakers should be designed with reasonably constant directivity and axial responses free of significant resonances, and that all headphone responses should be "Harman-like" in general insofar as having a similar shape for ear gain, and either defaulting to or at least having the option to have something close to the Harman target's overall tilt.
Like, a lot of Sean's commentary is mostly about headphones that...like, most people can buy, which are usually active DSP designs these days. Those definitely should have a "Harman preset", I think we can all agree that's a good idea. For passive headphones and audiophile DAC-amp stuff, having a robust DSP and measurements that would allow making profiles to EQ a headphone to be close to Harman sounds like a good idea, and I know Sean sees value in ex.
@oratory1990's database.
As far as how tightly grouped things need to be, at the very least when I've talked to Sean (both publicly and privately), he has seemed supportive of the idea of having options for different tastes, so long as they're like...real, identifiable-in-blind-listening-tests tastes, not just something somebody is asserting based on a sighted preference. I wouldn't really feel comfortable drawing boundaries for precisely what Sean or Floyd would consider "good enough", I guess that's something you'd have to ask them directly.
Re:
I would not say, and I don't think that Sean would say, that the publicly available Harman work has identified where ideal tone control adjustments would happen. Like, that just plain isn't something that was tested in the research, with the exception of allowing frequency adjustments on shelf filters. They may have done this internally, but in terms of what's published, we have results from a number of methodologies that all point in the same direction as far as what people prefer for the end result headphone frequency response, but not a lot of data to say which methodology is the best one for EQing.
It'd be a very interesting thing to do a paper on, in fact. Anyone got some spare budget?