There seems to be an easy fix for this on Google's end. When their automated systems see fraudulent activity on a site, instead of banning it entirely, they can

There seems to be an easy fix for this on Google's end. When their automated systems see fraudulent activity on a site, instead of banning it entirely, they can peg their future revenue to their historical revenue. This way, malicious publishers will have significantly limited incentive to engage in such behavior. In the short-term, their upside is heavily limited. And in the long-term, they risk being caught and banned. And at the same time, malicious blackmailers will also lose most of their leverage, like in the example given above. Even if the victim doesn't pay up, their income stream still remains mostly stable. And meanwhile, the attacker is spending a ton of resources on generating fake traffic that isn't earning them any money. This will eventually lead to the "business model" going extinct, which in itself solves the problem as well.