sure, but you didn't get the point. just because a model can predict certain results doesn't mean its correct. early astronomers were able to very precisely predict the positions (w/ looping motions and all) of stellar bodies in the night sky despite using a geocentric model. if you make it complex enough, you can engineer quite a few models that can predict natural phenomenon pretty well.
Yes of course, but this applies to any model. We cannot know of any model is "correct", they are just tools to describe what we observe. If they are more predictive than other models, then we should use them. Or course it's better to use a simpler model that gives the same predictions.
> if you make it complex enough, you can engineer quite a few models that can predict natural phenomenon pretty well.
Well, yes. Next thing is to pick the easiest of these models: Occam's razor. But pondering whether such model is "correct" won't bring any further improvements.
