The short answer
In principle, some supernaturalistic hypotheses probably can be scientific, including some standard claims made by creationists. In practice, however, creationists tend to treat their own claims in an unscientific manner.
The longer answer
I. Can supernaturalistic hypotheses be scientific in principle?
I actually agree that some supernaturalistic hypotheses can in principle be scientific. Although many scientists speak as though science by definition cannot say anything all about the supernatural, I do not think they are right. I also think most scientists would, on reflection, agree with some of the reasons I have to offer.
To start with, many supernaturalistic hypotheses are falsifiable because they entail observable consequences that can be checked. For instance, the hypothesis that God created the world 6,000 years ago entails the hypothesis that the world is 6,000 years old. Since the latter hypothesis is falsifiable, the former is falsifiable as well. In fact, the latter hypothesis has been falsified, so science has thus falsified young-earth creationism, as well as every other wider hypothesis that entails a young earth.
A trickier question is whether it is in principle possible for science to confirm a supernaturalistic hypothesis. It is, for instance, unclear whether evidence that the earth is 6,000 years old would constitute any evidence that a god, specifically, created the world 6,000 years ago. We cannot say that evidence that confirms a hypothesis also confirms any wider hypothesis that entails the first one, because there are infinitely many such wider hypotheses. The question really comes down to whether or not we can treat a god in the same way we treat theoretical entities like electrons; however, the legitimacy of a theoretical entity seems to be a matter of degree rather than a matter of kind.
Trying to figure out exactly what makes a theoretical entity legitimate will lead us into a thicket in philosophy of science that is best avoided here, so I will limit my comments to a few points. First, I agree with the standard view that most supernaturalistic hypotheses are too vague to be testable. For instance, since no testable prediction can be drawn from the general hypothesis that there is a god of some sort or another, that hypothesis is outside of the domain of science. It is quite true that science cannot address hypotheses like these, and I suspect that scientists who talk as though there is a strict line between science and supernaturalism have such general hypotheses in mind. Science and religion come into conflict when religion makes claims about observable characteristics of the world. But it seems to me that if one constrains a supernatural entity enough to specify precisely what one would expect to see if the entity existed, then such an entity could have the same role as any other theoretical entity. Of course, with sufficient constraints we might think we have just shunted the entity over from the realm of the supernatural into the realm of the natural, but I think that is just a matter of semantics. The important thing is which hypotheses are admissible into science, not what those hypotheses are called.
I think the history of science supports my view. For instance, scientists in Darwin’s time came to reject creationism not because it was supernaturalistic, but rather because they thought, correctly, that evolution did a much better job of explaining the data. Modern science presumes naturalism not because science innately excludes anything supernatural from consideration, but because supernaturalistic hypotheses have not stood the test of time, while naturalistic ones have done very well.
II. Are creationst hypotheses scientific in practice?
If what I have said above is right, then many supernaturalistic hypothesis can in principle be addressed by science. None of this says whether something like creationism is in practice at all scientific. The problem is that creationism is so malleable in the hands of its proponents that it ends up not making any predictions at all.
In the first place, their hypothesis often are so vague as to lack any real content, as when they offer the equivalent of “God did it somehow” as an explanation of the origin of the universe. Second, when they offer hypotheses that have enough content to be testable, they tend to make ad hoc appeals to the inscrutable will of God to square their hypotheses after the fact with disconfirming data. This kind of behavior strips creationism of any remote analogy to science. Again, the specific hypotheses advanced by creationists often are testable, and have in fact been shown to be false; it is the behavior of creationists, when presented with disconfirming data, that shows that what they are doing is not science.