> You simply can't get past what Gödel and Rice proved with current technology.

Not this again. Those theorems tell you nothing about your concerns. The worst case of a problem is not equal to its usual case.