The quotes from the post and the last comment are difficult to reconcile for me:
If we are in agreement on charging 0x03 validators for the opportunity to be selected to build the block, then it looks more viable. This is also where the pricing mechanism in attester–beacon proposer separation comes in. There is a fixed proportion of stake in the pool eligible for proposal rights, and the price adjusts dynamically to maintain this proportion.
If we believe that block-building rights will be resold anyway, it seems we can allow for a bigger pool of validators to participate, while they at the same time perform their consensus duties. And the cost of opting-in to the opportunity of being selected is priced according to the dynamic mechanism. However, since we may not feel the need to sell both proposal rights, it can indeed be only the execution proposal rights that are allocated in this manner, as this post suggests.
Let me know if I misunderstood something.