Suppose a person made a comment and a down-vote on another person's question. Why don't we have it so that when the other person makes the correction to his question, that the down vote will then be canceled (neutralized).
How do you propose to automatically test that the "correction" really did fix the reason the downvote was given?
How exactly would your mechanism work? Should it also cancel out upvotes? After all, the reason an upvote was given might also be destroyed by an edit. Should it remove comments from people who have voted? After all, if I downvote and add a "-1, you've made a mistake in this part right here" comment then surely that's confusing if it remains after the edit too.
There are a number of reason edits can happen. I might downvote an answer because of its poor overall quality - it might be poorly written AND contain factual errors. We have a lot of people on the site who enjoy editing for clarity and one of them might improve the readability of the answer but not change the problem of it being factually wrong.
Automatic cancellation of downvotes would remove my downvote at this point despite the answer still being factually wrong, and despite that being the main reason I cast a downvote. At this point your idea to automatically remove downvotes becomes dangerous; someone might take the information from that answer and use it. Lets hope the factual issues were not destructive ones eh?