Except that there’s no evidence that saying “killed” or “murdered” for example, actually does push you down the algorithm. And if it did, surely saying “unalived” would have the same effect by now, since the algorithm could easily be extended to this term, and it’s been in use for ages.
People censoring themselves in this way, preemptively, with no certainty that it is making a difference is like late-stage Orwellian batshit insanity.
It’s enough of a reason on its own to avoid those platforms. When even the creators are disingenuous enough to self-censor inoffensive words in an effort to appease an opaque algorithm, the content cannot possibly be meaningful.