One of the many things that SQLite brings is the ability to treat arbitrary storage as a relational database that can be queried (and mutated) with SQL.
Both of these things are generally rare in open source databases. There are very few multi-master distributed databases that have ACID transactions, and even fewer of them have SQL support. So marrying these two technologies is extremely compelling; you get the power of SQLite's relational data layer with the power of FoundationDB's distributed transactions. This is really high tech stuff for free!
A different implementation of similar semantics is rqlite. rqlite uses the Raft consensus protocol to replicate the text of your database writes to each node, which commits it to a local on-disk sqlite database. I haven't actually looked at mvSQLite enough to know how different the approach is, but both should be on everyone's list to take a look at and poke at. (For most "work" you'll be doing, it's probably completely fine to not create a distributed system like these; run in one datacenter, run a monolithic RDBMS, and tell users you're down when a tornado flattens the server racks. Do keep some offsite backups either way, though. Your application code will always have bugs, and these databases will happily replicate an accidental delete to 5 data centers at the speed of light.)