It just struck me that while it is very puzzling why there is law-like regularity at the bottom level—in fundamental physics—the puzzle about why there are law-like regularities at higher levels—in astronomy, psychology, biology, chemistry and non-basic physics—is a separate puzzle. In other words, even if we had an explanation of regularity at the bottom level, we would not thereby have an explanation of why there are higher explanatory levels where there are also regularities, albeit somewhat more approximate ones. Thus, when we are puzzled by the laws of nature, there are two things to be explained:
- Why there is regularity at the level of fundamental physics.
- Why this regularity, together with the initial conditions, gives rise to regularities at multiple higher levels of organization.
This gives rise to what one might call a generalized fine-tuning argument. The standard fine-tuning argument asks why the laws of nature (and especially the constants in them) are such that life arises. The generalized fine-tuning argument asks why it is that the laws of nature and initial conditions are such that multiple explanatory levels (either left unspecified like that, or enumerated: astronomy, psychology, biology, chemistry, etc.) arise from these laws and conditions.
Whether the generalized fine-tuning argument is good argument for the existence of God depends on two things: (a) how likely it is that apart from the theistic hypothesis that such multiple levels should arise, and (b) how likely it is on the theistic hypothesis that they should arise.
As for (b), I think in Aquinas and Leibniz we find compelling accounts of how an infinite but simple deity would have good reason to create a world that images his infinity via a diversity of elements and his simplicity via a unity running through these diverse elements. Unity at multiple explanatory levels allows even more of that diversity and unity.
What about (a)? I don't know. I think the question is easier when the levels are enumerated, as then the considerations from the standard fine-tuning arguments can be used. But the general question is quite interesting, too.