I like this direction a lot and have generally thought moving in the direction of eth1 as a pseudoshard of eth2 makes a lot of sense. Do you think there would be concerns about the eth1_friendly_validators system decreasing the current security of eth1? In general though, I Iike any approach that transitions eth1 into eth2 and lets it actively use the eth2 world.
My general thought is following Casey’s line of thinking, phase one and done. Adding basic execution into the shards should not be a heavy lift and we can establish a fee market EE/transfers EE. This could be used as the basis for operating the shards as a data availability layer while we work on innovating, securing, building and preparing more sophisticated contract/execution EEs.