Disagree profoundly.
And if perchance you got to that conclusion because you believe you can't apply the scientific method to CS (as in: formulate a theory about how the world "is", design an experiment to test the theory, run the experiment, rinse, repeat), you are mistaken: when dealing with systems of exploding complexity (which is, btw and imo, what physics is all about as well), this is a very fruitful attack strategy.
In my country, Computer Science is exclusively theoretical study of some areas of mathematics. You don't go into CS to run experiments, just like in mathematics. You go to study theorems and complexity theory.
Anything related to the act of developing software is called Software Engineering/Informatic Engineering. Totally different things: as different as a mathematician and a bridge builder. In the real world there's an evident intersection between CS and SE, but in academia (in my country), CS _is_ math.
Being able to predict how complex and long simulations will be is great as well.
BTW I think that "Computer Science" is interpreted differently in different countries. In France for instance we have _informatique_ which means "things with computers" and usually is understood as "development" or "system/network administration". _
As the académie put it: "science du traitement de l'information"
The method that you describe is used in exploring mathematical landscapes, including CS. But, still, some people (like me) find it useful to distinguish between mathematics (including CS) and science, by which we mean empirical science. But I would agree strongly that the boundary is very fuzzy.
