Speaking as a non-expert, my understanding is that the absence of fortifying data from the LHC was a major blow that demoted string theory because at least some of its major predictions in the heartland of its prediction space should have been accessible.

But that said, if string theory never makes an inch of progress on any question ever again, it will still have paid for itself in spades with AdS/CFT correspondence which has done real things, such as led to better predicted properties in quark-gluon plasma physics than alternatives, and of course, serving as the basis for black hole holography and from there, as a potential road map to holographic universe. People grew a little too comfortable waving away string theory as completely untestable and I just don't think it's fair to say that anymore.

And now, and most importantly, string theory is at least the grandparent of the current New Big Thing, which is generalized flat space holography, taking holography from black holes and seeing if there's a version of it that applies to the rest of spacetime. There is real excitement there, and it's a topic that, baffingly to me at least, still awaits it's Brian Greene style public champion. If generalized holography ever ends up being worked out into a theory and validated, it would qualify as a major conceptual revolution in physics. We're not anywhere there but even in the absence of string theory as a great unifier, there seem to be tantalizing convergences between quantum error correction and holography that suggest some big important thing they might share in common.

Again I'm not an expert and barely know what I'm talking about, just curious enough to read about it, but I feel like a lot of the string theory excitement from yesteryear should be transferring over to holography right about now.