I think only really good outcome research is going to tell.
I know people who began such practices, and have since ceased them.
It's a damn good thing here, with our remote communities, to do crisis assessments via tele-link for things like whether someone needs to be hospitalized and, over time to do remote diagnosis for serious mental illness, since mental health professionals are only getting anywhere near some communities every eight or nine weeks.
Both clients and professionals say they find it better than nothing, but that it still sucks.
I know Monash University did a research program on running online anxiety diagnosis and treatment recently. I note that this has ceased...but I have not seen the research ourcomes yet. They will be interesting, as anxiety really responds a lot to cognitive/behavioural input, and ought to be a good candidate for online methods for those with mild to moderate symptoms.
I can't imagine working that way a lot of the time...but I certainly do a lot of between session phone contact with remote clients, and do a lot of supporting other professionals with their clients in that way.
I know there was a lot of talk at some stage about setting up sophisticated online robot programs to work with people...and I saw some reports that some people preferred that to actually seeing someone.
I haven't seen research on that.
Actually, that sort of group stuff ( I just read the article) is likely to be pretty damn helpful...especially for folk disinclined or unable to do it live.
It kinda happens here.
I'm worried about someone who needs to talk about bloody "syndromes" ("wife-abandonment syndrome" yet!) in order to talk about perfectly reasonable grief, though.
Syndrome inventors fill me with syndrome-inventor dread syndrome.