Please note before I start this post that I'm an atheist and don't literally believe any specific religion is true.
I can start an arbitrary religion that states any belief in a god leads to eternal suffering... So don't believe in god, and lose nothing... believe in god and lose everything. Why is the religion I just invented less likely to be true than others?
I think it's fair to say, even as an atheist, that there's a non-0 chance that some god exists, and a slightly lower chance, but still non-0, that some specific existing religion is true.
If we limit the discussion to Christianity or Islam, for example, to simplify the conversation, then we would have to conclude that IF there's a non-0 chance that these religions are true, then there's a non-0 chance that their God is as they describe in some sense, and that God seems to want people to believe in him, and also show himself or communicate with a select few individuals, so he's made some amount of evidence of himself existing (very weak evidence) in the world, in the form of religious writings.
Now, it's my opinion that the probability of one of these gods existing is FAR higher than the probability of a God you've just made up existing. These holy books are weak evidence, but they are evidence, and therefore they put these specific conceptions of God more probably than a conception of God that you've just created for the sake of a thought experiment.
I think that's possibly why this isn't a very popular argument against Pascal's wager - I think people realize that any idea you invent about God just for the sake of a philosophical argument is naturally going to be lower-probability than a specific existing popular religion.