Mike2, your issue is epistemological.
In general, the lifetime of a theory in physics is built in the following path:
1/-You make a few observations in the physical world. These observations may not be clearly related to each other in a straightforward way.
2/-You think of a theory (equations and so on...) to explain them all. This step is among the hardest.
3/-Your theory successfully accounts for phenomena observed in 1/ (here is what you called "the curve fitting process")
4/-Skyrocketed by such a success, you, or another physicist, predict NEW ODD THINGS from your theory, things that nobody has ever thought of nor seen before.
5/-Experimentalists, willingly or by chance (rarely now), find out that these NEW ODD THINGS actually exist and are in total agreement with what your theory predicted.
6/-Unfortunately, some people, working night and day, later find out a new phenomenon that can't definitely be embraced by your theory. You are stunned but this is life.
7/- A young talented physicist publishes an outstanding article in which he brilliantly demonstrates that your theory was a special occurrence of HIS new theory when, say, velocity is low compared to "c".
Right now, quantum mechanics is at stage 5/. It seems likely that we won't reach stage 7/ before a few decades. Newton's physics were overthrown by Einstein in 1905.
The point is: building a theory in physics is a long process which is the result of interactions between theory and observation. You cannot just throw observations into the garbage and expect to build quite a new theory throughout a thought process only. Such a way is of doing physics is not fertile and the history of science has proven it.
