
This was an interesting article revealing a bit of infighting among physicists wrt string theory.
IMHO, it points to an ideological difference which would stem from the priority given to pure mathematics in physics. Indeed we've seen similar disputes here on the forum between those of us who center on the mathematics (information theory, complexity, etc.) related to evolution and those who center on the sciences (biology, chemistry, genetics, paleontology) .
It is a philosophical difference which I believe we would all benefit from exploring.
Personally, I fall on the “math first” side of the debate – I put mathematics above all sciences  and physics at the top of the science heap because of its integration with the mathematics.
My reasoning is that mathematical structures are universals per se and thus their discovery sheds light on all the sciences. As an example I assert Einstein’s ability to pull Riemannian geometry offtheshelf to describe relativity. Other examples include dualities and mirror images  the "unreasonable effectiveness of math".
An example related the subject of this article  some string theorists like Vafa (my personal favorite) treat the physical ramifications of their work as secondary, e.g. he was not troubled by the consequence of an additional time dimension in fTheory though it caused quite a stir in the physics community at large. IOW, good mathematics may not be widely accepted or have an application at the moment, but will remain "on the shelf"  because they are universals, which may be relevant at some other opportunity.
Additionally, even if physicists were not pursing a "theory of everything"  the mathematicians would. As an example, here's a link to Jurgen Schmidhuber's Algorithmic Theory of Everything.
I’d love to hear any arguments for why mathematics should not be given a higher seat in our body of knowledge than science!
One last point  I do agree with everyone here who instinctively believe that mathematical truths ought to be conceptually simple. The KaluzaKlein based compactified string theories are not without competition, e.g. the easy to comprehend noncompactified 5D higher dimensional dynamics.