Recently, I voted to keep a question open, believing that it was answerable and valid in its current form. I then answered the question, by posting a terse and incomplete answer, deleting it, and then taking the time to edit the deleted answer to answer more fully. As it turned out, by the time I'd completed writing the full version of my answer, the question had indeed been closed. But because my initial incomplete answer had been posted while the question was open, I was able to edit and repost my answer in its full form.
While I'm proud of the answer I posted (and still personally think the particular question was valid and appropriate for the stack) I found a point Thomas Markov made elsewhere quite compelling:
The question's author returned and accepted the answer after the question was closed, and now has no reason at all to resolve the issues with the question.
This is an excellent point. The closure procedure is there to help improve questions, and maintain the integrity of the stack. The fact that users cannot answer a closed question is key to that process, incentivizing the community to edit and improve the question if they'd like to answer, and encouraging the asker of the question to improve it in order to get useful and helpful answers. But the ability to edit an existing answer allows users who answered quickly to circumvent this process. This could lead to questions which would otherwise have been edited and improved being left in their sub-optimal state, after receiving an answer they find useful.
So my question is this: should steps be taken to ensure that a closed question not only locks out new answers, but also prevents users from un-deleting and/or editing answers they've already posted on that question until the question is reopened?