The mechanical analogy is particularly interesting here because at least one of the claimed vulnerabilities involves tricking the victim into decrypting an encrypted message for the attacker and then sending it to them. If someone can be tricked into opening a safe to let the burgler rummage around inside then few would consider that a failure of the safe technology. I mean there is still a problem there but it is a different one.

I think this supports my contention that we spend much too much time quibbling about cryptographic trivialities when it comes to end to end encrypted messaging. We should spend more time on the usability of such systems.