Thank you for the write-up!
What do you think of the risk that the scheme incentivises investment into fast-connections to proposers, such that the bid can be made just-in-time and particularly after most other bids were committed? This seems especially true for large proposers who may colocate with builders, and thus earn outsized rewards compared to their stake, defeating smaller proposers more often. Generally I feel there needs to be a stronger argument for why we expect proposers to win with a frequency proportional to their fraction of the total stake.
One thought that comes to mind is why wouldn’t all eligible proposers also attempt to burn the block reward? It looks like the dollar auction example (without the costly second bid), where a $1 bill is put up for sale. In that case, bidders would be willing to pay up to $1 to receive the item. So if block reward is 0.05 ETH, I am proposer A and I observe MEV = 100 ETH, and my bid is 100, then proposer B can best-reply with a bid equal to 100 + 0.01. Sure there is a cost, but B still wins the auction and receives 0.04 ETH. At the limit, shouldn’t it be expected that all eligible proposers would bid away the block reward too?