[fauigerzigerk]

I don't think "social media" is anywhere well defined enough to compare its effects with the link between cigarette smoke and cancer.

It's far more difficult to prove specific effects of social media use, because it's so hard to separate it from all other social and economic dynamics of our time. There's no proper control group.

The way in which different people use social media may well make all the difference between it being beneficial and being harmful. Social media is not a substance that has a specific effect regardless of how we use it as individuals and as a society.

I think it's a mistake to expect social sciences to work like natural sciences.

The smoking gun you're hoping for may never be found. The entire question could end up in the same basket as worries about the alleged harmfulness of reading novels or comic books, watching TV, watching porn or playing computer games. If anything "social media" seems even more vague than all of these.