>So that means all bets are off: it's entirely possible to implement a perfect peer reviewed protocol in a client and/or server full of backdoors that entirely

>So that means all bets are off: it's entirely possible to implement a perfect peer reviewed protocol in a client and/or server full of backdoors that entirely bypasses the protocol deliberately? No, that's not how cryptography works. Signal clients don't trust the servers, and it's entirely possible to review the client protocol and make sure that malicious server can't learn anything about the communication (except the dreaded metadata - who talks to whom). That's the point of E2E encryption.