Plain and simple.. should string theory be considered science or philosophy?
It's a scientific theory...
I say that because a well-known young German string theorist named Urs Schreiber is now in the Math Department of his university (not in the physics department)
and an even more famous older American string theorist named Michael Douglas just posted an article that he wrote in the archive section for "Differential Geometry" (which is mathematics) rather than the usual "Hep-th" (high-energy-physics-theory) section where he has posted papers in the past.
Another young researcher Aaron Bergman, known to many of us, recently wrote a propos of something "...that's why I'm doing mostly mathematics now." I forget what the discussion was about.
Penguin, you might want to think about what empirical science is----where the theories are something you can test, and potentially falsify, by observation and experiment.
And compare that with mathematics, where one traditionally does not do experiments. One sets up axioms which dont have to be true. then one sees what can be logically derived from those axioms by ASSUMING that those axioms are true.
so to take a artificial example, one can have a 5 dimensional world shaped like a donut and one then proves things about that world----facts that follow logically from its being 5 dimensional and in some precise specific way "donut-like".
this is the axiomatic method, not the experimental method. it is not EMPIRICAL SCIENCE but it has often turned out to be extremely interesting and sometimes even useful.
Separate names with a comma.