Not opposed. I, and a number of others, have been tossing similar ideas around recently. Especially in light of the complexities of a generally extensible and feature rich Phase 2 being pretty massive.
Bringing eth1 natively into eth2 soon 
will allow for a more native use of the eth2 data layer for massive layer-2 (1.5?) scalability in the meantime without the induced delays and complexities that emerge in a finality-gadget + bridge mechanism. This would also be a natural place for introducing validator liquidity without an emergent second asset.
In this path, eth1 could still become a single-shard EE if/when EE functionality is built and deployed. As long as we do some due diligence on the structure/framing, we can still evolve toward the promise of a general and extensible Phase 2.