This is only a response to your edit:
EDIT: From what I experienced, experiments are already producing results, when theorists are still trying to fit their theories to the data. So why do you need the theoretical calculations then? Do they have predictive power that would be found easier and more precise with experiments?
What sort of predictive power can you get from experiments? Experiments only let you "predict" something by actually carrying it out. That's neither a prediction nor a retrodiction--you could call it 'diction', I guess ;).
If you carry out multiple experiments and use their results to predict stuff, you are theorizing about the nature of physics. That means, you have a theory. if you want to make predictions with experiments, then a theory is unavoidable. On the other hand, experiments can make retrodictions--basically verifying a theory with experimental results.
The issue is, while we try to make a general theory based on experimental results, more results keep coming in. Leads to a bit of an issue when the new results don't fit in. Of course, there's the flipside fanfare when they do fit in (Prediction of gallium, prediction of $\Omega^-$, Gravitational lensing--and if the Higgs is found, we will have quite a bit of fanfare)
Here's an extremely simple analogy(lifted from a math.SE post), which may explain the reason why theories never can keep up with experiments: In my experiment, I take natural numbers from $1,2,3,4...100$ and compare them with $10^6$. I discover the exotic property that all of them are less than $10^6$. From this, I theorize that all natural numbers are smaller than $10^6$. I feel happy in having creating a theory that checks out with experiments. The theory has use in the everyday world as well--we don't deal with such large numbers anyways. Now, someone decides to test this theory further. He tries larger numbers (no doubt using a Large Number Collider with floating-point arithmetic), and discovers that my theory no longer holds.
Note that my theory is still pretty applicable, if someone asks me "how much money is in your pocket?", I can safely answer "less than a million" without having to count the money or know how much is there. But, if I did deal with that kind of money, my theory would no longer hold. Similar things happen in physics. Experiments rule out old theories, but they simultaneously set bounds for which they are valid. Theory comes from a half-baked perception of the world(imagine if I gave you a slice of a car and told you to figure out how it works), which is why it must keep up with experiments.