Yes, it's just like all high level studies trying to make broad statements about human behavior or any other complex human system, like nutrition or economics. The reason there is so much contradiction is that we can't deal with the complex relationships between variables or the unknown unknowns. The only reason medicine is able to do decent studies is because they can tightly control things in RCTs. Correlational/observational studies are almost never good enough to make strong statements about anything.

It might make more sense to say that social media usage is just one large aspect of a child- like the presence or lack of presence of supportive adult figures, the presence or lack of presence of food security, good education, physical safety, etc.

I'm sure a teenager who runs exclusively in cutting or pro-anorexia social circles online is a much different teenager to one who primarily uses the internet to keep up to date on horse shows.

You got one too eh? tw sh / tw ed?

I don't know what this means; I don't have children if that's what you're asking.

I think the "sh" and "ed" decode to "self-harm" and "eating disorder". The "tw" might be "trigger warning". Given that, the "you got one too?" might mean "you have an eating disorder too?", although it doesn't make grammatical sense as "you have a self-harm too?", although that could be interpreted as "you have a self-harm disorder too?". But I don't really know.

I believe they're tracking related symptoms and disagreeing about the root cause.