I discovered a simple typo that is pretty prevalent. Can this be fixed with the "link" replacement tool, given that it's actually "a simple Find and Replace tool"? (That linked Q&A is very related, but didn't make it clear if non-links specifically can be replaced.)
And if so, is there a lower limit to how many instances there would need to be in order to get something like this approved? In this specific case, there are 100+ posts with misspellings (and 500+ posts with proper spellings, possibly with overlap) on a single site. The site is also not Stack Overflow, so I would have to edit a few posts every day for the next month plus in order to fix it without causing any problems. That's really hard to commit to, even for me. (In fact, I think I've been aware of this specific problem for over a year, maybe multiple years.)
The following patterns should match almost all the typos, except for a few that have unusual punctuation bordering it (underscore represents space, and "word" is a placeholder):
_Word__Word,_Word._word__word,_word.
This probably wouldn't break anything, which I could verify more rigorously with SEDE beforehand. (And if it was going to break a few posts, it would be easier to revert those changes than to do everything manually.)
I know that there are community-run projects out there that automate edits. However, these still bump the post and therefore need to rate limit themselves. If you were to spend some time looking for suitable typos to automate away, I'm certain you'd find enough to keep such a bot busy for years. Hence this question.
