I usually see a lot of suggested edits pending, suggesting to remove a user sign or cheers from a question.
Es.
//question ...
thanks
Mark
What's the right policy, approve or reject?
If that's the only wart of the post, approve. If there's much more wrong with it, reject as too minor¹. If there's a tiny speck² besides, improve.
That's my policy for these suggested edits.
¹ I try to go and fix it myself afterward, but I don't always succeed with that.
² Don't take the "tiny speck" too literally, it's for dramatic effect.
Approve.
Salutations should be removed - they are noise that adds nothing to the question.
See Should 'Hi', 'thanks,' taglines, and salutations be removed from posts?
I agree with Daniel's answer, but I now apply a stricter policy when I see users making significant edit suggestions in a manner that I think is primarily to gain reputation.
If I don't recognize the editor's name, then I'm more likely to hit the Improve button and fix things myself. If I recognize the editor's name as likely doing the edits specifically to gain reputation, I'm far more likely to hit the Reject button and describe why I rejected the edit: it's the easiest mechanism for me to provide feedback on how future edits will have the most value. After all, there's probably another 400+ edits coming until the user hits the +1000 reputation cap from edits.
For example, a new editor has suggested 75 edits in two or three days. Hooray! Except his editing is sometimes at random, sometimes right-on but insufficient, and sometimes he gets it exactly right.
I'm glad this user is choosing to make so many edits, but some of them appear like he searched for an awkward phrase ("see the code behind") to replace it with something more idiomatic ("see the above code") and ignores the other glaring issues in the posts.
He started with simple edits. They've improved over time; I think the improvement is largely because 11 of his 75 suggestions were rejected with useful reasons.