You cannot defend what you cannot precisely define and threat model. Non-secure platforms cannot be defended nor threat modeled. If you send a military/intellig

You cannot defend what you cannot precisely define and threat model. Non-secure platforms cannot be defended nor threat modeled. If you send a military/intelligence critical message to someone then you need E2E encryption. In order to asses correctness you need more people to buy in than those only backing military applications. This is found in technologies like TOR which wouldn't be effective it couldn't be used by non-military agents. If there is no need to include the greater community, than there is no pressure on sponsored developers to pursue mathematical correctness.