Your Marilyn pages are an excellent source of mental workout and entertainment. I wanted to join the discussion of Pascal's wager. Religion is a touchy subject and I have labored to approach the discussion within the constraints of the wager.
Pascal's argument reduces religion to a gamble with an investment and a possible payoff; it was specifically designed this way in order to appeal to a wide audience of various beliefs. I therefore contend that it isn't specifically belief in God that Pascal advocates, but the practice of belief. As you point out, belief itself is not so easily won or lost. His rationale is much more oriented towards convincing people to be good and to pray and so forth, with the potential of a future, eternal, reward.
In the wager's basic form, the investment is assumed to be zero, so that the expected return, no matter what chance of winning there was, would be greater than the investment.
However, the practice of belief, as Cynthia points out, may entail a certain positive investment -- in the form of relinquishing one's vices, spending time practicing the religion, etc. The value of this investment, of course, varies from person to person.
Without that positive investment, Marilyn does have some validity when she says that the best bet is to join whatever religion makes the most promises. But that can easily be shot down if we consider the investment; usually the religions with the most promises require the most effort from your earthly life. These would be the high-risk, high-return investments of the religious realm. To decide on a particular religion using the math of Pascal's wager (where actual belief is irrelevant) requires a careful assessment of how much you will give up for that religion, weighed against expected return of the religion. This is similar to a studio deciding which movie to produce, or an individual choosing between a bacon cheeseburger and a garden salad.
Pascal was pretty smart. He probably thought of all this. There does exist a second level to his argument that accounts for the above reasoning. Any positive amount of effort you invest in the religion will be finite, because earthly life is finite. But if the gamble pays off, the reward you get is eternal, thus there is infinite return. When you multiply the return by its odds of occurring, assuming those odds are nonzero, you have calculated the expected return, and it's still infinite. Therefore, it is still true that religion is always a good bet!
I'm trying to think of the atheist's response most compatible with that logic. The soundest approach I can think of is this. The law of diminishing returns would apply to an eternal life, so that the value of that eternal life would actually be finite. The total utility of an eternal life, then, would be comparable to the utility of earthly life.
Interested in your response. I'm also gonna see if Marilyn can come up with anything.