To what extent will the complexity of Ethereum 2.0's base layer bleed into app development?
This seems like one of the big open questions remaining on Ethereum's path to world domination.
I'm aware of a few different types of layer 2 efforts, as well as the new OVM research which provides a unified theory and base layer API for all layer 2 systems. These are really exciting and progress is huge! But it occurs to me these are building blocks and not a specific answer to the question of "how will Eth 2 app developers actually deal with shards?"
A related open question I have is around state rent. What happens if Compound Finance forgets to pay its state rent and a major part of DeFi infrastructure gets deleted? This problem seems easier than cross-shard transactions. A simple example solution might be a DeFi app paying upfront for a "10 year lease" on storage.
TL;DR what do we imagine is the best case scenario for Eth 2 devs dealing with shards?