As I mentioned before, Jason Zarri has an interesting post entitled, How Free is God's Will? in which he makes the argument that God's existence can only be an explanation if God's will is not completely free. In effect, the argument goes like this:
(1a) Suppose that God's existence is compatible with possible worlds which are disordered and hostile to life.
(1b) Then the existence of God cannot explain the order and life-friendliness of the universe because those characteristics are no more likely if God exists than if He doesn’t.
(2a) Suppose that God's existence is incompatible with possible worlds which are disordered and hostile to life.
(2b) Then there are significant constraints on what God can will, for then God cannot actualize just any possible world.
From which it follows that either God is completely free, in which case God's existence does not explain order and life-friendliness in the universe, or God's existence does explain order and life-friendliness in the universe, in which case God is not completely free.
I think the argument is falling victim to an ambiguity about what is meant when we say, "Y is explained by X's existence." On the one hand, we could mean that Y is explained by the bare fact of X's existence: from the proposition "X exists" we can directly infer (either defeasibly or indefeasibly) "Y exists (or occurs)." However, if we think about how we talk about causal explanation, we virtually never talk about causal explanations in such a way as to mean this. If I say, "The trash is gone because there is a trash collector," I am not saying that from "This person that is the trash collector exists" I can infer "The trash is gone." I am saying instead that "The trash is gone" can be explained given that there is a trash collector who takes away the trash; the one explains the other because from "The trash is gone" I can reasonably conclude "The trash collector did the sort of thing that makes it to be the case that the trash be gone." What explains here is not mere fact of existence but the fact that the existent thing is engaging in a particular kind of action; in other words, what explains is not simply that something exists but that it exists in a particular causal role.
Likewise here. When someone says, "The existence of order in the universe can be explained by God's existence," they aren't saying that you can infer anything from God's existence, but that you can explain the explanandum if there is something that exists in the right kind of explaining role, e.g., God. Thus the advocate of the teleological argument is saying that the order and life-friendliness of the universe can be explained if something, namely God, does a particular sort of thing, which requires, of course, that God exist, not that the order and life-friendliness of the universe can be explained simply by an appeal to God's existence.
Jason is exactly right that if the advocate of the teleological argument tries to explain the order and life-friendliness of the universe purely in terms of God's existence, and not in terms of God's being the only sort of thing that can plausibly fill the sort of causal role that is required for explanation, that he is committed to saying that God is not completely free. In fact, the only way that will work is if the effects of order and life-friendliness follow necessarily from the mere fact of God's existence; only when we can assume that effects are related to their causes deterministically can we infer (indefeasibly) the effects from the mere existence of their causes. This is an entirely general point. Suppose a detective is trying to solve a case and suddenly realizes that the facts can only be explained if Moriarty exists. The only way he can take the mere existence of Moriarty to explain the facts is if he thinks that the facts follow merely from Moriarty's existence, without any free choices by Moriarty intervening. And things remain much the same if we are talking about defeasible inferences: for the mere existence of Moriarty to explain the facts, the mere existence of Moriarty has to rule out possibilities relevant to the facts, and thus the very existence of Moriarty implies limits on Moriarty's freedom.
But usually we would expect the detective to mean that the facts can only be explained if a certain causal role is filled (if certain things have been done) and that Moriarty is the only plausible candidate for fulfilling that role (and thus must exist in order to fufill it). Thus, while Jason is right that God cannot be completely free if effects like 'order' and 'life-friendliness' follow from God's existence with either probability or certainty, nonetheless this does not generalize, because most causal explanations are not explanations by appeals to mere existence but by appeals to causal roles that imply (defeasibly or indefeasibly) the explananda if something exists to serve in that role. When people say that God's existence is necessary to explain certain facts, they are saying that the facts can only obtain if certain things are done that only God can do; since the facts obtain, those things must be done, and thus God must do them; and in order to do them God must exist. Thus the explananda are not explained in terms of the possibilities available to their cause, but entirely in terms of the necessary conditions of the explananda themselves. There might be no possible world, even a hostile world, incompatible with God's existence; what matters for the explanation, however, is that the existence of an orderly and life-friendly world is incompatible with the non-fulfillment of a causal role only God can fulfill. That's a rather different sort of incompatibility. And mutatis mutandis if we weaken the connections here so that we are dealing with probabilities rather than necessities.