In my experience as a suggested edit reviewer on Stack Overflow, a vast majority of suggested edits are about formatting and grammar. Only a small fraction require subject knowledge in the first place.
When you see a suggested edit that requires subject knowledge that you don't have, click skip. It's a very simple rule: if you don't know, click skip.
Sadly, there is a part of the Stack Overflow community that systematically rejects corrections to code in answers, in direct contradiction with the editing guidelines (“When should I edit posts? (…) To correct minor mistakes”). They suggest commenting instead, in direct contradiction with the commenting guidelines (“When shouldn't I comment? (…) Suggesting corrections that don't fundamentally change the meaning of the post; instead, make or suggest an edit”). I'm trying to educate them.
Requiring subject knowledge from the reviewers would not improve things. The problem is the attitude that rejects corrections, not the lack of subject knowledge by the reviewers. Requiring score in a question's tags would make reviewing take longer, some suggested edits in smaller tags could easily sit around for months. It would not improve anything regarding correct edits getting rejected, and it would only slightly reduce the time wasted by reviewers who have to skip a review because they lack the knowledge to vote on it.