A fined tuned argument.....Heywood style.

[Alex K]

* Even if I accept all your premises, we don't get 50% each by a looong shot. This is because for those two propositions, observations very strongly favor B over A: The Universe looks exactly as you would expect an anthropically selected universe to look: if you have a priori probability distributions with strong dependences on the parameters, you expect to have many parameters of your theory pushed towards catastrophic boundaries, but barely on the anthropically favored side. Here's a published paper where this is explained:

http://arxiv.org/pdf/1005.2783.pdf

Alex, I don't find this critique meaningful because any universe would look exactly as you would expect a universe to look if B were true. This artifact of B isn't sufficient cause to favor B over A.

The only reason to favor B over A is if somehow you concluded objectively that the universe does not look like A. I don't think you can come to that conclusion objectively. Just because you would have designed the universe differently isn't sufficient reason to dismiss A. What you find elegant and beautiful , another intellect might not. If you designed the universe you might make the cosmological constant exactly 0. If I designed the universe, I might make the cosmological constant very close to 0....but not 0.

If you set aside your bias and personal feelings about what a designed universe should look like, there is no objective reason to favor B or A.

Ok, you just systematically removed your hypothesis from testability because you proclaim that it does not make any predictions as to how the universe looks, and you think this somehow strengthens your hypothesis, but it doesn't. If hypothesis A makes no prediction for a certain observation, but hypothesis B does, and observations match the predictions of hypothesis B, then B is to be favored over A at least with respect to these data.