George Orwell, 1984:
Newspeak is the only language in the world whose vocabulary gets smaller every year.
Now, you are actually proposing that we use Newspeak: a language that's based on English, but from which some words are deliberately dropped - in hope that
- if it can't be named, it doesn't exist, or at least
- restricting the language makes for easier control.
Even worse, the list of dropped words is completely arbitrary, as the words are innocuous in most contexts, and the list itself is secret.
So, taken ad absurdum, let's imagine that the people who make the filtering software will get offended by the word
code tomorrow (as it's so obviously used on those perverted internets!), and decide to include it in their naughty-word-list. Will you come here asking for that word to be removed from all questions? But this is the same problem that exists with the term "adult"! There's no guarantee that the list won't be updated tomorrow to also block "grown-up", or "breasts", or "saggy", or "homosexual"; also, there is no One List To Rule Them All - should we keep anything which might potentially in some situation possibly maybe offend one person off the network? That's as good as shutting it down.
There's another problem to self-censorship: it wastes enormous resources. "Is this word likely to be banned? And this one? And this one? How about if I replace it with that one?" Asking for such additional effort will push people away.
Also, there are no true synonyms: by editing the post, you are changing its meaning without necessarily improving it. "Grown-up" sounds like the kids' version of the adult term "adult" - you've already shifted the meaning slightly. The sentiment around here is "edit iff it improves the post and doesn't alter its meaning," such "anti-censoring" edits would likely fail one or both of these criteria, and you might be seen as editing for editing's sake.
TL;DR: No. Basic decency is already required for the SE network; ridiculously convoluted self-censorship to appease a filter would be sacrificing too much for a tiny and uncertain gain.