That would still be relative though- if things improved in 6 weeks, the worse could be the first two or three, if things improved in a few years, the worse could be the first years.

Also, those few studies are measuring certain types of limited outcomes in relatively short term studies, so I'm not sure they're too applicable to a bigger picture analysis of this, perhaps to folks who go with CBT or something, for specific concerns.

Also, depending on the term, clients could be enjoying the honeymoon phase of therapy where the relief at getting help and having someone to unload on feels like an improvement, but is not all there's going to be in terms of the work or progress which often comes later.

I get a bit concerned about study generalizations when links to the studies or specific detail isn't provided. A link to a meta-analysis is even better.

I think people entering therapy, virtually by definition, have something that isn't working... and facing that and learning how to change often entails difficulties, otherwise, a simpler, more common intervention would've worked.

P.S. I don't think worse before better must be taken as an all-or-nothing experience. For example, a sense of lonliness or alienation may decrease upon confiding in a therapist while a sense of disillusionment about the struggle that brought one there may increase as we discuss it, just one possible permutation. I think another adage, that change is stressful, applies, even with good change!