I am perfectly willing to stake FreeBSD's security on the belief that nobody will ever perform a brute-force search against a 256-bit keyspace.
For that matter, Microsoft, Apple, RedHat, Debian, and Ubuntu all make the same or (usually) even weaker assumptions in their respective packaging and/or updating systems.
It's an idea roughly analogous to SRP, which is well-studied. Share a key, then later prove you still know it.
Client-side hashing would be great, if there was a way to do it that didn't involve Javascript.
Your proposal also treats a 2-way cryptographic function as a 1-way cryptographic function. That is, optimistically, a questionable thing to do.