The interesting bit is the meta question implied by this - whether truths developed in a mathematical sense are valid in other contexts.
AN answer is something along the lines of this:
While a single equation cannot be created to fit every possible model it IS possible to develop an equation that fits properties of the model under study (at least to your own level of understanding of both maths and the problem domain).
The question whether mathematical insight can be used as an analogy machine to determine outcomes in other domains is the same question as to the breath of any particular philosophy, IMO.
To come back down to the question at hand Consider that the proposition under study is that differing eras of paradims are incommesurable.
Given the new meta framework we can then ask what would the underlying scientific model changes between physics, biology, and the social sciences be that would necessarily invalidate this proposition?
Since all of these models ultimately rest in mathematical descriptions of experimentation on created models, the question appears to me to be moot. That is, given the basis of these disiplines they cannot help fall into the same category. Even broadening their functions to the philosophical does not lead one out of this conclusion (if one accepts Maths as simply A particular rigorous philosophy).