Even in write-heavy environments, I've used SQLite effectively with multiple thousands of concurrent (as in logged in and changing things at the speed of human

Even in write-heavy environments, I've used SQLite effectively with multiple thousands of concurrent (as in logged in and changing things at the speed of human site interactions, not as in simultaneous web requests at the same instant) users. In this context, SQLite is effectively a write linearizer. With appropriate pragmas and database handles stored persistently between requests in server processes, the rate (in bytes of data affected by INSERT/UPDATE/DELETE) at which users could make DB changes got pretty darn close to the rate at which a file could be sequentially written on the filesystem, which is more than enough aggregate throughput for 1000s of concurrent non-bulk changes in most web applications, I'd imagine.