A very suble reasoning.
I'm confident everybody has fully understood it but just to make sure: would you please explain?
Anyways, it doesnt lead nowhere. The assertion "If God is conceived of as an unobservable transcendental being, then one could not disprove his existence" is equivalent to "If God is..., then one could not prove his non-existance". Which means "God does not exist" is not verifiable.
So, where are we now?
I can't put it any simpler than this: a falsifiable statement is something that it is impossible to prove right but is possible to prove wrong. Negative statements are necessarily falsifiable, in that you can prove them wrong by proving a counter-example.
Verifiability and falsifiability are different things.
In the current context, Billie -Joe suggests we should all be agnostic about God (but not about faeries or unicorns, for reasons he has yet to logically explain). However, if we pose the God hypothesis in a falsifiable form - "God does not exist" - then instead of agnosticism, we are necessarily lead to consider that, given the lack of evidence *for* God, it is sensible to believe more strongly that he doesn't exist.
Dawkins talks about this at some length in The God Delusion, but the thrust of his argument is precisely upon these lines - just because something might be possible (yes, God *might* exist), the probability of it occurring is not equal to it not occurring.
Science "believes" that, for example, the speed of light is a constant. This is a falsifiable, testable hypothesis, and as such there always remains the possibility that it might not be. The evidence, of course, points in the opposite direction, and thus it is sensible and rational to conclude that the speed of light is indeed a constant. We can never say we have "proven" that the speed of light is *always* a constant, but the evidence allows us to preclude taking an agnostic position.
Falsifiability is resolutely NOT about remaining on the fence. It is about producing more certain, and not less certain, hypotheses. As long as you know what will prove your theory false, you can always know what evidence makes your hypothesis stronger (though never "proven correct").