This was very interesting, stem, thank you for laying out Barnes' argument. As a probability argument, he never seems to admit that positing a supernatural god should have a probability far less than the low probability of our universe, which because it actually does exist and is observable has a greater than zero probability. An observable event like that should always be assigned a greater probability than the idea of a god, which is given with no comparable observable evidence.
Eta it seems honorenthos is making a similar argument, but from the standpoint of the low probability of the evolving universe still not excluding an actual event; my argument is that comparing a probability between an observed event and a postulated event (god) the observed event by definition always should be assigned the higher probability, and therefore considered more likely.
I know people who use fine tuning as their ultimate reason for being a theist. I just don't know how you get to a Christian/Muslim/Hindu... god from fine tuning. It seems to appear many times with "look at the trees" while ignoring the Loa loa eye boring worm. And what guarantee is there that the god is good? Maybe he/she/it is evil.