There's plenty of prior art to comb through if you're actually asking how difficult it can be? this working group at a linux foundation org where i worked interviewed representatives of every current and recent form of prior art they cd find...for 2 yrs!
https://github.com/decentralized-identity/edv-use-cases

This is pretty interesting as far as I understand it. Tackling data storage puzzles figuring out the tangled bits, connecting different specs, and making sense out of it.

What’s the trajectory here for protocol labs do you think?

lots of related problemspace exploration in the protocol labs space as well, honestly. I also worked for a company for a while called Spruce Systems that did some great work on these problems as well