Fair point. Information technology needs refreshing every 3-9 years or so (Moore's Law and such, though Moore treads more lightly over algorithms than he does microchips).
People ... need a years to develop, and could potentially provide useful service for decades. Having an entire cohort of agents removed is a substantial blow.
U.S. reliance on technology is so beyond reasonable at this point that, I'd imagine there's the very real, serious liklihood of something on the level of an Enigma/Lorenz cypher botch that gives away the whole country on a platter.
It's something that I've wondered about for a while now, but really, who am I to question such things?
That's an interesting and scary idea, but keep in mind that a lot has happened since WWII. Back then industrial-mechanical cryptography was seen as this magic wand that made all communications completely inscrutable. Then the Polish and British and others went to such crazy lengths building brute-forcing machines, etc, to actually break these ciphers, and since then the idea of an infallible crypto system is out of style.
So I think the strategy is much different and much more durable now. The military designs ciphers to be resistant to theoretical future quantum computer architectures. They fund research every year to prove various properties of crypto systems. And I'll bet there are lots of fallbacks for any critical infrastructure.
People ... need a years to develop, and could potentially provide useful service for decades. Having an entire cohort of agents removed is a substantial blow.