That’s an interesting thought, and could have some merit. Two issues I see with it - the first is that all instances would have to be on board. It’d have to be accepted and used across all Lemmy instances as a core Lemmy feature for it to be effective. The second is that some of those bad votes already fall into reporting territory. I don’t know what the reporting system is like on desktop but on my app it’s only a text box to report. Reddit had what I thought was a good system that was preset reports for site-wide and then subreddit specific rules. I think a robust reporting system (and again, maybe there is one or there’s one in the works) could handle spam, troll, or bot posts.