The perceived poverty of supernatural explanations

When it comes to arguments for design, in favour of some form of theism, the arguments tend to concentrate on areas where the is scant empirical evidence, where the science is speculative.
Examples of this sort are theistic hypothesis for the origin of the universe, for the supposed fine tuning of the free parameters of the universe and for the origin of life on Earth. It seems that people believe that wherever there is a gap in our knowledge, there is justification for pushing God into it.

As far as I can tell, however, these arguments fail, for a number of reasons.
"God hypothesis" in general seem to lack empirical content. They are literally explanations for any possible empirical evidence. If you're unable to confirm or falsify an hypothesis, then you have no means of knowing whether it is (provisionally) correct or incorrect. Proponents of such "God hypothesis" tend to retreat into mystery, claiming that since God is outside of the "world", then we cannot expect to find "scientific evidence" of it's existence or actions. Perhaps this is correct, and I'm willing to entertain the notion, at least for arguments sake. This does not get the theist off the hook, however.

As argued at length by Greg Dawes in "Theism and Explanation", a "God hypothesis" is an intentional explanation, rather than the sort of mechanistic explanation we usually find in science. Rather than giving a "God hypothesis" a free pass, however, this simply requires a different sort of explanation be tendered. The proponent of the "God hypothesis" ought to supply detail of the intentions of the putative agent, it's character, it's goals and desires, it's capabilities. From this detail the phenomena to explained should follow, such as why we find ourselves in a material universe, why the "fine tuned" parameters have the specific values they do, or why life was started on this particular planet, using the particular processes we see around us. To avoid an ad-hoc explanation, the "God hypothesis" should also provide further "predictions" of empirical observations we should find which also follow from the intentional explanation as provided. These would allow us to test the explanation.

Providing an explanation of the sort Dawes argues is required, would constrain the "God" of the hypothesis in, I'm sure, many ways. And it is an open question as to whether a possibly successful intentional explanation of this sort could or would be accepted by a theist. The deity, so constrained by explanatory requirements, may not match up with a theistic believers "internal" beliefs about who or what God is.

So, we have good reason for not accepting the simple God hypothesis which are generally presented by theists.

I think, apart from the explanatory problems with "God hypothesis" as they currently seem to stand, there is also good reason in many cases, to expect that a naturalistic explanation is preferable, even where there is little to no empirical support for such an hypothesis.

A naturalistic explanation would not require the expansion of our ontology to include "immaterial minds". If we were to have solid external reasons to accept the existence of God, then perhaps the explanation would become preferable, but at present this does not seem to be the case. Natural explanations, even when speculative, tend to be extensions of existing scientific knowledge (origin of life research relies upon biochemistry, fine tuning explanations rely upon existing physics models, as do origin of the universe explanations). The science that these models are based upon seem to be solidly attested, and while the speculative hypothesis may extend the science in ways which are not yet evidenced, they tend to do so cautiously. Naturalistic (scientific) explanations are also usually formulated to be open to empirical confirmation/falsification.

Another reason to prefer a naturalistic explanation is simply the lessons of history. For every phenomena which has a satisfactory explanation, that explanation is a naturalistic (scientific) one. Claiming that a current gap in our knowledge is likely or certain to have a supernatural explanation is like betting on a horse has lost every race it has entered. It is to put yourself in a position comparable to a believer in Zeus, Thor or Yahweh some centuries or millennia ago, claiming that because there is no alternative explanation for lightning, it simply must be whatever god the believer had faith in.