I don't think that's quite the problem. In mathematics, the word "theory" is often used when referring to particular mathematical frameworks (e.g. Group Theory, Graph Theory, Morse Theory). In that sense I think String Theory is very much a theory. As you imply, in physics, the word "theory" is typically used in a different sense. I'm not a physicist but I presume a physical theory has to be verifiable, consistent with observations, able to predict the behavior of unexplained phenomena. If I understand correctly, the basic problem is that in some quarters string theory is being passed off as a physical theory. I know of pure mathematicians who are interested in string theory and who couldn't care less whether its a physical theory.