My closest colleague has tried to pin them down: string theorists assert with near clockwork regularity that string theory is 'perturbatively finite', but when you ask them where this is proven, most can't point to a source, and those who can point to a paper in which the methods only work up to two loops. . .or they give bogus arguments based on misunderstandings of a mathematical construct called Teichmuller space. And, to date string theory can only be done in a 'flat background' space-- but general relativity tells us gravity is equivalent to certain kinds of curvature on space. If it's not perturbatively finite, or can't be done in a realistic space-time, string theory isn't even a theory much less a theory of everything.
Of course, neither the popular science press nor the National Science Foundation seem to have noticed this, not even when Edward Witten, one of the leading lights of string theory over its 20 year life has given up on it to turn to work on loop quantum gravity--which according to the string-theorist quoted in the times (who has no basis in either physical or mathematical fact for his assertion) 'must be part of string theory'. It isn't.
Huh?
BTW, for some TOE technobabble BS fun in a sci-fi way, read the book called "Distress", by Greg Egan. Takes this stuff and twists it around into ultimately a silly conclusion, but it was kinda fun reading it.. (the person who "discovers" or first interprets the TOE becomes the "keystone", or something like the creator of the Universe....)