A Curious Move in Robin Collins' Defense of the Fine-Tuning Argument

I recently skimmed Robin Collins' chapter on the fine-tuning version of the design argument in The Rationality of Theism (Ed. Paul Copan). I had read it before about a year or so ago, but I didn't catch a problem I noticed this time around. The worry is this. In a section defending the argument against the "Who Designed the Designer" criticism, Collins uses Swinburne's reply that, roughly, an explanatory posit y can explain some phenomenon x even if y is itself complex-yet-unexplained. However, when Collins discusses the "Many Universes" criticism of the fine-tuning argument, he argues that such a hypothesis is implausible, on the grounds that the mechanism that would be required to produce the universes on that hypothesis is complex and functional, and thus would itself require a designer.

In short, Collins seems to accept the following principle when he responds to the "who designed the designer?" criticism of the design argument:

(*) A theoretical posit y can be an adequate explanation of some phenomenon x, even if y is itself complex-yet-unexplained.

However, when the critic uses the Many Universes hypothesis to explain the data of fine-tuning, Collins responds in a way that is in conflict with (*).

So the worry is this. There seems to be no principled way to qualify (*) in a way that makes it legitimate to apply to theism, yet illegitimate to apply to the many universes hypothesis. But if not, then he must either accept (*) in its current, unqualified form or reject it. Now if he rejects it, then unless he takes God to be simple in structure, his hypothesis of a theistic Fine-Tuner can't be a legitimate explanation of cosmic fine-tuning, and thus his design argument collapses.

So he can't take that route. So he must accept (*). But if so, then he loses the reply to the Many Universes hypothesis, and with it any explanatory upper-hand the theistic hypothesis may have had with respect to the data of fine-tuning.

Of course, he could reply that God is simple -- e.g., he could hold to the Thomistic conception of God as a being whose existence is identical with his essence, and that all of his attributes are identical. But he seems to be hesitant to do so in his chapter (and rightly so. See, e.g., Plantinga's Does God Have a Nature? for the problems with such a view). Still, I guess that one could say that, contrary to anything we can imagine, God is absolutely simple, and our inability to comprehend God's simplicity is due to our cognitive limitations. But at that point, the pretence of offering an explanatory account of fine-tuning disappears; and in any case the many-universe proponent could play that game: perhaps at the most fundamental level of reality in the multiverse, reality is ultimately simple.