The Parsimonious Multiverse?

A number of people (physicists, other scientists, and interested laypersons) have prominently taken some big swings at string theory recently. (I saw at least two books on a recent walk through a local Barnes & Noble.) I'll comment on string theory another time, but here's what troubles me: If we're demanding empirical results, what about the widely discussed anthropic principle and "many worlds" interpretation of quantum mechanics? A couple of articles have recently referred to the anthropic principle, arguing from opposite sides in the theism/atheism debate.

Richard Dawkins uses the anthropic principle it in a recent essay, which I assume is in sync with his recently published book, to address questions about cosmology, while a cogent, well-written piece by (presumed) theist Frederick Turner in another forum uses the consequences of the anthropic principle against atheism (in a limited sense).

But is the anthropic principle truly parsimonious? What physical evidence, if any, actually supports the idea that the observable universe represents but a mote in an (essentially) infinite sea of possible universes (or "multiverse")? One of the better arguments for the multiverse I have read was the quantum computing argument from David Deutsch. But in the end, his argument is explicitly about the explanatory power of the multiverse theory, NOT about anything that is required by the data, as far as I can tell. That is, pure quantum math and reasoning produces the results of quantum computation, and nothing about the multiverse makes any prediction(s) that are distinct from "standard" "uncontroversial" quantum analysis. If all the multiverse does is to provide a convenient resolution to some metaphysical questions that trouble some (many?) cosmologists and quantum philosophers, then surely it is less than science and even worse than string theory in this regard. It must somehow produce predictions as well as have explanatory power, if it is going to count as real science. Frederick Turner points out that the multiverse thesis would indicate that God (or something indistinguishable from God) should exist in at least some subset of the multiverse! This is, of course, far from demonstrating that our universe has a God, but a clever bit of logical jujitsu nonetheless. (Bravo, Mr. Turner! But I really must balk at the wildly overreaching assertion that religion is "perhaps the most powerful… survival strategy.") But isn't this "proof" that God is out there somewhere (if not necessarily here) really an excellent illustration that the whole multiverse/anthropic principle meta-cosmology is NOT parsimonious? If God in our own universe is not parsimonious (and I tend to agree with the atheistic notion that He/She is not) then a bunch of alternate universes with their own "Gods" hardly seems elegant and Occam-slashed.