Your concern is understandable. Edits that change the post without improving it waste the time of reviewers and people viewing the front page. If the edit fixes all the obvious issues with the post, it should be approved.
First, the review task in question.
In this case, I would probably have approved that edit if it tried to fix the topicality. If there was no topicality issue, there'd be nothing wrong with approving that edit - it makes the post and site better by removing cruft. The more we can show people we're not a standard forum, the better. Note that the Improve Edit button is good for cases where the editor makes a good effort toward a polished post but just missed a thing or two. If the question weren't off-topic, I would use that button to adjust the grammar of the first couple sentences. Alas, it is off-topic, so rejecting the edit is understandable.
When it comes to alerting users of edit rejections, there's already a thing for that: a notice will appear when the edit suggester goes to edit another post. The warning gets more and more severe as more rejections accumulate. If the edit was approved despite your reject vote, one option is to talk to the user in chat (if they use it) and help them do better next time.
If you're wondering how you stack up with other reviewers in terms of accept-to-total-reviews ratio, I wrote an SEDE query that orders all Suggested Edits reviewers from most permissive (i.e. most likely to use the Accept button) to least. You come in as the 14th least permissive out of the 342 reviewers who looked at >20 suggestions. You use Approve on 59.39% of the reviews you perform. Considering the quality of some edit suggestions, I think that's just fine.
(Note that this query completely ignores "Improve Edit" and "Reject and Edit" votes; they're not in the
SuggestedEditVotes table and there's no
UserId column in
ReviewTaskResults. There are MSE requests to change the former and the latter.)