The whole thing is designed to allow them to call people (especially women) "mid". They arbitrarily chose to use a gaussian distribution pattern so they had an excuse not to give anyone a score beyond 6s.
But real answer is be a hugely successful fashion model who the sub creator found attractive. Then their "objective rating standards" would include arbitrary criteria to bundle your face in. The whole sub could be replaced with a trivial ML model if it were actually about just giving their "objective" ratings. The internal weights used by the ML model would make about as much sense as the crap spouted in the screenshot.
There is the tiniest, infinitesimal amount of value in the statement that, likely because of the way we’re all graded in school, we don’t really use the full range of a 1-10 scale for attractiveness, and are sorta only really saying ugly, mediocre, attractive, or model.
There is zero value in saying that that’s an issue and the solution is a psychopathic and dehumanizing system.
When it comes to looks, the only meaningful scale consist of a 0 (not attractive) and 1 (attractive). E.g., if a woman I am talking to doesn't find me attractive, there is zero practical difference between me looking like Chris Hemsworth or an ogre.