There can be Bayesian evidence for non-falsifyable hypotheses. You might perhaps be interested in "Belief in the Implied Invisible "
You know the drill - If it's worth saying, but not worth its own post (even in Discussion), then it goes here.