Take the "creepyness" part, which is also a multiplace function of the beholder and beholden and context, and you've got the same problem.

I guess I shouldn't have assumed it was obvious that I was scope-masking both "creepy" and "attractive" under respective "as perceived by whoever is making the attractiveness/creepyness judgment at the time where this parameter is relevant" formulas.

So, to factor, unpack, inline and reiterate: Ceteris paribus, actions or behaviors or phrases always appear more creepy to a given observer or participant whenever said observer or participant finds the source of the actions, behaviors or phrases less attractive at the time of evaluation where the level of creepyness is evaluated by said observer or participant, and conversely appear less creepy when the source is found more attractive under the same circumstances.

To me, by charitable reading when taking LessWrong as context, the above paragraph and the first one in the grandparent seem equivalent. Should I not be reading others' comments like this mentally? I've been doing this on every comment I read for months.