Consider it this way: natively speaking English means you have domain knowledge, sharing that domain knowledge should feel no different from pointing out mistakes in other domains. If someone gets English wrong, that's no different from getting, say, VAT calculation wrong.

Maybe I'm just assuming everyone already handles mistakes in other domains with tact and respect.

Also, it's a matter of proportion. If a function name is extremely misleading that's different from a function name just using an awkward synonym, just as a broken piece of code is different from dodgy indentation. You'll want to fix either problem eventually, but the immediate repercussions are quite different (serious bug vs minor annoyance, basically).

I think we're generally in agreement – I was trying to keep the focus on things which can be reasonably be considered objectively wrong and keeping the degree of wrongness in mind when deciding how to handle it. If you're on the low end of that scale, it's probably not a good use for everyone's time in a code-review session unless it happens a lot.