It looks like a parallel implementation of TiDB, which they had built in Go. I would assume that avoiding garbage collection and reducing heap usage were valid motivations for the project. But it doesn't really have enough information on the repo to tell what their motivations are.

It's transactional and mvcc, you can't say the same for most KVS around.

To be fair, I was considering learning Rust and implementing a distributed hash table just as an exercise. With a like-minded team, reimplementing a toy Spanner sounds like a fun project.

Language is one of the more important features for attracting other contributors to open source projects.

As to reinventing the wheel, given the number of competing tire companies that exist apparently even the wheel still has room to improve.