I’ve been thinking about this problem a bit more, and it seems like even in the existing system validators can informally fix the price of gas by voting down the gas limit until congestion occurs, or even by simply forming a 51% cartel that refuses to accept transactions with less gas than the fixed price or mine on top of / vote for blocks that include cheaper transactions. With enough incentive I don’t see any reason why this wouldn’t happen, as Ethereum miners already coordinate things like the gas limit fairly efficiently.
I am increasingly suspecting that in any system where validators earn most of their revenue through transaction fees and coordination costs between validators aren’t extremely high, miner cartels raising transaction fees is going to happen. Validator cartels probably have a lower social cost than a “tragedy of the commons” scenario where there isn’t enough demand for auction-based transaction fees to cover miner costs anyway, so perhaps a voted-on parameter deciding transaction fees might not be as bad as it sounds, as it gives a “legitimate” channel for monopoly pricing.
But again, the fact that @vbuterin never mentioned it his quite comprehensive paper on transaction fee pricing systems probably means it’s garbage. I’d really like a rigorous breakdown of why though.