* Because the therapy is asking the patient to do something that, at some level they don't want to do and think is going to be painful, therapists generally have to exert significant effort to persuade the patient to continue to the point where they benefit.
* Most (all?) Psychological therapies have a high failure rate. A significant fraction of the patients won't be helped, and so the persuasion above is harming a large fraction of patients (the opportunity cost of a failed therapy, which generally takes months, is huge)
I have not seen this problem acknowledged at all by the psychological professions. There ought to be a good deal of research into how to decide the optimal stopping point, and making sure that the operation of a therapy causes the least possible harm to those is doesn't help. But my impression is that the cognitive dissonance of delivering an intervention that will help some, but expensively fail many, is too much for most therapists. They generally continue to believe that the patient would be helped if they held on for long enough, and don't provide any support for a rational process for deciding when to give up on a given approach, resulting in the kind of gaslighting and blaming described here.