Indeed, you make the important point that the theories we choose are indeed choices, and that is always true in every scientific paper ever written. This I feel is a very important and common misconception about science-- many people think that the theories of science are handed to us, and we have no choice in them. But that isn't how it works in practice-- everyone who ever wrote a scientific analysis of some phenomenon always started out, in every case, by making an array of choices in how they would do that. So it is undeniably clear that the act of doing science gives an important role to making choices, that are not completely arbitrary, but are flexible to the goals at hand, as you say.

Toward that end, I think it is quite important to avoid language that tends to further the misconception that any aspect of science is devoid of such choices. So I think we should be careful when saying "If you believe that science's job is to tell us what the Universe really does", because I would say that anyone who believes that is much more like a creationist than they would care to admit-- in that they are holding to beliefs that are patently refuted by scads of evidence to the contrary. But I think I know what you intended by that, which perhaps is closer to "If you believe the ultimate goal of science is to expand our understanding, and create models that are in some sense 'closer and closer' to what is really happening." If that's our goal, we will never be content with a theory that we know is less accurate than some other one. But still the first step is doing science always pretty much boils down to "choose your poison." And as someone once said, "it's the dose that makes the poison", so I don't mean that this is in any way a futile or unproductive choice to have to make.