That doesn't sound like an underlying problem that sounds like a coincidence and like greedy idjits doing the thing in the wrong way

Like yeah, content that doesn't call for genocide is also more advertiser friendly in addition to being safer for marginalized communities in the real world.

I think the solution is a moderating format that involves more open source work, assign accounts a value that weighs their reports by how serious the accusation is, but also by how often they've turned up credible reports in the past, have the report come with three parts, the reported post or comment, the rule or rules the reporter think it violates, and a text box for them to give some explanation of why they think it breaks the rules in case it isn't obvious.

Once someone has a high enough rating and volume for reports, earmark them to become a paid moderator.

Und voilà, you have home grown community moderators familiar with the rules and with sneaky ways bad actors try to get around them, along with a process to quickly get more whenever you launch in a new community.