"Personally, I think the problem will be solved by a new conceptual approach,"
As a computer scientist, I have to admit to being partial to more discrete theories like loop quantum gravity. But whereas I'm willing to stick my nose into some science's business and declare they're all Doing It Wrong, this is not one of them. String theory is often criticized for having a lot of predictions and not much evidence, but I think that to be fair, that's true of all the competition, too. There just isn't enough evidence right now.
I think it's likely that relativity and quantum mechanics will be reconciled by an information-centric theory--which could easily come out of computer science.
We call some patterns of energy "information" (like an algorithm) and other patterns of energy "matter" (like a proton). It seems to me that the recent mathematical study of information could be extended to cover matter as well.
Information theory is also a much younger branch of study than physics, which would seem to indicate a longer runway for fundamental discoveries and applications.
As a computer scientist, I have to admit to being partial to more discrete theories like loop quantum gravity. But whereas I'm willing to stick my nose into some science's business and declare they're all Doing It Wrong, this is not one of them. String theory is often criticized for having a lot of predictions and not much evidence, but I think that to be fair, that's true of all the competition, too. There just isn't enough evidence right now.