The problem is, the borders upon which you want a child operating in social media are pretty fuzzy. Do you want them working with classmates? sure. Other peers in same school grade? of course; older peers? ...sure; older peers in other districts...maybe?
Then there's all the spoofing and the "age gate" software that inevitably needs to be done to do this.
The things is, it’s not the people that bother me. If it was IRC or MySpace I’d be mostly fine, even if they were engaging with questionable people or content, I think partially because the fidelity and partially because those experiences were still largely pulled by the child. It’s the non-stop algorithmic content.
This is sort of - it’s not really the “social media” that’s the problem it’s billion dollar companies getting to push content direct.