My experience with concolic testing has been that single path satisfiability queries tend to be quite cheap and it's actually a bit challenging to write the too

My experience with concolic testing has been that single path satisfiability queries tend to be quite cheap and it's actually a bit challenging to write the tooling around the solver in such a way that the solver would actually become the bottleneck. This is not to say that combinations of concolic testing with other testing methods is not an interesting research topic :)