I'm no expert on the IKKT model, so don't know exactly what the problems with it are. If it really did give a well-defined, background-independent, non-perturbative formulation of M-theory, one that seemed to lead to dynamically determining a plausible background, my guess is that everyone in the field would be working on it. As far as I can tell, most people have lost interest in it. When that happens there's generally a good reason. Unfortunately I doubt that killing the IKKT model would kill the whole string/M-theory field.
I'm not claiming string theory is completely uninteresting. I am claiming it is overhyped and it has become difficult for young people to make careers working on anything else, which is very unhealthy. The reasons I don't believe in string/M-theory 11-d supersymmetric unification are two-fold:
1. It's really complicated and ugly
2. It not only doesn't predict your phone number, it doesn't predict anything at all.