You haven't proven that there are better designs out there being kept down by the priesthood, just waved your hands in the direction of Grand Conspiracy and anointed Bruce Schneier a member in that supposed priesthood. The thing is, cryptography isn't kept obscure. It's a mathematical discipline welded to the practical discipline of actually writing software, but so are other topics in that part of CS, such as compression and formal verification.

But here's the nub of my point: It is hard, though, and it's harder than it looks, and this matters more for cryptography than compression because of the things people do with cryptography. If you design a bum compression algorithm, you make files you can't really decompress and you get discouraged. If you make a bum encryption algorithm, and you get all excited and actually use it for something important, and it's easy to break, well, you might get a lot worse than discouraged. Major companies have been bitten by this. Is your algorithm better than HDCP? Probably not. You know what happened to HDCP? Yeah. Intel wishes it had taken the kind of advice any new cryptographer gets about not taking their own inventions too seriously.