"Dark matter" reminds me of a humorous thing I once read about:
Flanaghan's Finangling Factor: The number which you need to multiply your answer by to get the answer you should have gotten.
Science and engineering has more of these than we would like to admit.
Well, that's for sure, on the other hand, you have to have at least some numbers to start with - you can't derive any number out of "nowhere", you need to start from something; e.g. all non-quantum physics can be built upon the value of speed of light and charge of electron - all the rest comes out of the theory...
Lets not get this confused with the numbers that "magically" appear in equations - physicists do use phrases like "nature likes pretty numbers", but that's not reason most equations contain numbers like 2, 1/2, 1.732... (square root of 3), etc.; all these "pretty" numbers come out of integration - dimension of space is whole number after all, ain't it?
Then again, the quantum theory (esp. the supersymmetry and string theories) sure have quite a few "parameters" - e.g. supersymmetry in most formulations has about 10 parameters; yet, lets not forget that these are typically "situated" in more dimensions, thus we need more "start numbers" to get to our universe, with just fewer dimensions...
The thing about dark matter is that you can't see it or touch it and nobody can tell you what it is. It's there to make the equations work.
That's true, and to many people (and some 10 years ago also to many scientists) this was a big problem; but lets not forget that e.g. electrons and protons as such do not exist also (and I'm not talking only about the particle-wave duality) - they're simply a mathematical concept that is easier to grasp if presented as an existing "particle" (as wonderfully described in
Surely You're Joking, Mr. Feynman! when he asks philosophers what does the "abstract concept" mean)... the same goes for
all particles of the Standard Model,
quarks, photos, etc. ...