Well since a common thing therapists do is (covert) social skills training, increased extroversion makes perfect sense. The was recently another one linked here that said long term therapy can increase neuroticism.

My biggest complaint is that most of their studies suffer from survivor bias. They ignore people who drop out or have negative outcomes (there is overlap there).

My biggest complaint is studies being inadequately placebo-controlled. It's not just that studies often have no control group or a blatantly different control condition (waitlist, pill placebo), it's that there's no consensus on how psychotherapy works. That means there's no agreed-upon approach for sham psychotherapy, and even assertions that it's impossible to do because virtually any therapy-like experience already contains most of the beneficial elements (cf. [1]).

[1] https://en.wikipedia.org/wiki/Common_factors_theory

Yes, why is the article implying there is something philosophically bad about a changed personality? Ideally, if someone was a serial killer, it would be objectively good to change their personality wouldn't it?

Changing one's personality doesn't take away 'me-ness' anymore than it reveals your ideal self.

Imo no personality is ever wrong, except for when the brain/mind is influenced by a disease. I am not saying that the actions of a serial killer will ever be approved, but the notion that one "type" of personality is more good than any other is strictly cultural.

At the same time, I agree with you that it is not bad to change a personality, if it is the owner of that personality that initiate the change.