It's a known fact that users who've cast a lot of votes qualify for potentially having their votes preserved in case they request their account be deleted, because invalidating all their votes as is done normally can cause too much disruption to a lot of other users. In that case, an employee has to go through the profile and check to see if there's any reason not to preserve the votes (e.g. voting fraud) before making a decision on whether to preserve votes (transfer them to the Community user) or invalidate them as normal.
Recently, I've come across two documented cases (one here, can't find the link to the other) where, after waiting a full week after the deletion timer said their account would be deleted, the account still hasn't been deleted. It would be nice to have more insight into the background process behind this procedure, including how often the queue of users awaiting vote inspection for deletion is checked, and how long it takes for the actual investigation.
Also, what if a user cancels the deletion during this delay? Will the deletion be cancelled, or will it still be carried out since they let the original timer lapse after 24 hours?
Finally, if a user who's deletion is "stuck in the mud" contacts SE, will it expedite the request or will it just hold it up further? It used to be that hold-ups for vote inspection would cause the deletion notice to temporarily disappear (as if it were cancelled), and re-requesting deletion would prolong the process.