I think you really missed the point of my post. It was a hypothetical “random person” scenario. Literally any person walking down the street. My point was regardless of their physical demeanour: “effeminate” or “masculine” or “reserved” or “flamboyant” or “pick whatever adjective you like” - there is no way to accurately infer that person’s sexuality. In a general context there is no neccesary connection.
I wasn’t saying anything about monitoring individuals in real life for the sake of guessing their sexuality or other business. God knows, I spent 90 % of my time in the city, all cities I’ve visited, being that lone and lonely (or self/absorbed) guy. You are picking a beef where there is none to be had.
However, that said, if you change the context from ‘random observation of random person on random street at random time’ to something more correlatable with sexuality then the correlation between observable demeanour and sexual orientation might be stronger. But then the real determinant is location.
I have in mind Oxford st Sydney at 9 PM on a saturday night. The majority of people there are likely gay but they’ll present across a glorious range of demeanours that make the idea of an essential gayness that can be spotted pretty ridiculous.