Are there any ideas that's potentially more useful than implementing sharding?

Justin, have you quantified how much different features could improve Ethereum’s scalability, or of storage and I/O, computation and bandwidth? Where can I find more information for all of this research? For instance, if I search for “Merkle Mountain Ranges” in https://github.com/ethereum/research, I get no results. Maybe it’s in the private repo: https://github.com/ethresearch/main, that I and the public don’t have access to? (If I could get access that’d be great.) I remember seeing stuff on storage rent (which would still be useful with stateless clients e.g. for archival purposes), and I know there are articles by Christian Reitwessner and Vitalik on SNARKS & STARKs, which I’m still reading through. I’ve also seen data availability posts on this site, as well as stateless clients. I haven’t read anything about a lot of the topics that you listed for storage and I/O.