"Compatible" blindspots may make an alliance feel smoother, but ultimately it impedes progress, so it should become apparent over time.

The only issue I would have with this writer is the idea that a generalization is invalid if there exist any exceptions. All generalizations have to be scrutinized, but just because they don't apply in every circumstance doesn't necessarily mean they have no usefulness. And I think he leaves the impression that theoretical knowledge is more of a hindrance than a help, and I would disagree with that. Such knowledge needs to inform and infuse the process of therapy, but not be in control of the process. That's why I think therapy is an Art supported by Science, rather than a Science practiced artfully.

ETA: Although he doesn't address it directly, his flexibility hypothesis also supports the idea of benign boundary crossings as potentially pivotal moments of progress--and there has been a certain amount of research indicating the value of such moments of flexibility. His approach would seem to give more prominence to such moments as a frame.