Strategy specific builders: to censor or not to censor

Opening this issue to start discussion around the topic of strategy censoring

Front-running is illegal in traditional financial markets so it is no stretch of the imagination to think that < large validator set run by corporate entity > [ Coinbase, etc ] may be required to enforce a similar restriction on its validator's MEV augmented profits.

Therefore, some type of strategy specific censorship is inevitable. And even if regulation isn't the driving force, it could be possible to see more altruistic focused validator sets coming from community governed groups such as Lido.

Some open questions around this:
- How to form open standards around types of arbitrage?
- Can this choice be made in protocol (builder, relay, validator?) or is it best left to be socially coordinated?
- Can we accurately classify MEV types?
- If a heuristic based classification model is used, what are the risks of misclassification?
- Will ignoring certain types of MEV lead searchers to construct builders for their censored strategy, and would this be a bad thing?

The issue names no files, tests, or implementation entry points; it opens a discussion about censorship standards and whether choices belong in the builder, relay, or validator. Start by reviewing those protocol roles and the unanswered questions, but a concrete definition of done is not provided.

Written by the indexing model from the issue text.