An interesting reverse approach: instead of waiting for good tokens to displace bad tokens, it's better to work on the incentive mechanism.

Rather than viewing data farming and data refreshing as the protocol's "persistent problems," it might be more productive to see them as part of the participation process—by designing reasonable mechanisms, turning those studios and groups that once profited from bulk operations into "shareholders" or "core contributors" of the protocol.

The key lies in changing the revenue model: they no longer earn through token arbitrage like cutting leeks, but through long-term actual gains tied to the protocol's growth. When participants' interests are linked to the health of the protocol, the incentive structure reverses—those who manipulate data start to genuinely care about the product itself, because false prosperity will ultimately backfire on their own gains.

Can this redesign break the dilemma of "revenue from using the product vs. revenue from data manipulation"? It's worth exploring.