The linked study is interesting and useful and I'm surprised I wasn't aware of it. Thanks for posting.
Still, these sorts of things are difficult to interpret because severity of problems are also correlated with length of therapy and inversely correlated with outcome. So it's hard to interpret, even if you try to statistically control for it.
It's also something that would be difficult to randomly assign because different lengths of therapy might not be a good match for someone.
It's a bit like saying "length of rehab following traumatic accidents aren't correlated with as positive of outcomes". That might or might not be true but if it were it would be difficult to interpret because you could imagine rehab would take longer after serious accidents and also have more difficult outcomes.
There's other things to consider with length, like maybe the focus is shifting? Maybe someone continues to benefit from it?
Anyway, I agree it's important to try to remain focused in therapy and not use it in the wrong ways, but length is a difficult thing to make sense of sometimes.
We do have randomised trials to support the argument, but (like most treatments) there's a very obvious dose-response curve. If we take outcome measures after each therapy session, we can see that each additional therapy session has a diminishing marginal benefit.
There's a reasonable argument that the most severely affected patients might gain clinically meaningful benefits from very small incremental gains, but that's very difficult to falsify because the effect size is so small. Long-term psychotherapy undoubtedly fails any reasonable cost-benefit test, but even setting that aside, I'd suggest that anyone who isn't satisfied with the results of therapy after 12-18 sessions would be better off trying something else - either a different therapist, a different psychotherapeutic modality, or a biomedical treatment.
Anecdotally, the vast majority of people who I encounter who are in long-term psychotherapy aren't severely unwell; obviously that is in large part explained by the barriers to access faced by most severely unwell patients. Many of these people aren't "mentally ill" in any meaningful sense, exhibiting only subclinical symptoms of anxiety or depression even at the point of starting therapy. They don't have clear goals for therapy, or any real sense of when it would make sense for them to finish therapy. If they find therapy to be an enjoyable activity like doing yoga or having a massage then I have no qualms with that, but many of them ascribe benefits to long-term therapy that simply aren't supported by the evidence.
Still, these sorts of things are difficult to interpret because severity of problems are also correlated with length of therapy and inversely correlated with outcome. So it's hard to interpret, even if you try to statistically control for it.
It's also something that would be difficult to randomly assign because different lengths of therapy might not be a good match for someone.
It's a bit like saying "length of rehab following traumatic accidents aren't correlated with as positive of outcomes". That might or might not be true but if it were it would be difficult to interpret because you could imagine rehab would take longer after serious accidents and also have more difficult outcomes.
There's other things to consider with length, like maybe the focus is shifting? Maybe someone continues to benefit from it?
Anyway, I agree it's important to try to remain focused in therapy and not use it in the wrong ways, but length is a difficult thing to make sense of sometimes.