Pools would prefer to rid themselves of the regulatory risk of block templating… but not at the cost of reduced profitability for their customers.  So they need some guarantees that miners will create good templates, or a way to penalize them if they don’t.

The reality is SV2 still allows the pool an opportunity to veto templates from miners anyway.  So if pressured by authorities, they could still censor enough to be “OFAC compliant”. 

 StratumV2 is better than stratumv1 for sure, but fundamentally the pool can always reject the block templates. I think it would be good to crowdsource block template generation, and Demand pool is pushing the boundary in that regard. Giving miners the responsibility means the state needs to go after more people to enforce censorship, which can make Bitcoin more censorship resistant.

But still if the pool can reject the block template or withhold payment, they’re still trusted third parties. I don’t think this is a bad thing though. Big trustless pools open themselves up to block withholding attacks, but small trusted pools have less of an attack surface. More small trusted pools across more jurisdictions means miners have more choice, and means the state must attack more pools and miners if it tries to censor Bitcoin.

Anything that we can do to make Bitcoin more resilient should be done.