To be even more specific (https://www.sciencedirect.com/science/article/abs/pii/S01401...) heavy use of social media by children has been shown to correlate with reduced life satisfaction, but moderate use has no correlation with any effect on children. There isn't even any good research showing causation, just correlation, so prohibitions on children using social media will have absolutely no effect on most children, and for the rest there's a possibly of a benefit but also a possibility that there is no benefit, and a symptom of something larger is no longer visible.

That's a pretty small gain, in exchange for requiring everyone to give in to constant surveillance.