But we both know that your fanciful hypothetical is not how that tool will be used in over 99.999% of cases, because that kind of thing almost never happens, while moderator abuse often does.
Just because there are cases where a tool can be used well doesn't mean that introducing it in to a system rampant with corruption and abuse won't make things worse.
The hypothetical you chose is disgusting, unnecessary, and impossible in the real world.
Arguments to child porn I think are the modern equivalent of godwin's law - the moment it gets brought up, all productive discussion stops and is replaced by emotion.
Plus, moderators already use bots to instantly pre-ban things things and then go through and manually whitelist them.
The distinction is between banning for posting vs banning for upvoting. One reasonable argument I can see is that Reddit goes to some length to hide what you up and down vote in every other context, and people might be able to use the ban tool to dox people who thought votes, as opposed to submissions, were wholely anonymous.