If you implemented that simple solution the expected outcome is businesses collecting ID at the door. But unlike the age verification bills there'd be no prohibition of or penalty for misuse of the collected information. It's a strictly worse outcome.
You can make intentional targeting illegal without criminalizing the accidental. And mandating self categorization of content by service providers would enable standardized filtering that was broadly effective.
The above won't get kids off of social media and it won't serve the purposes of the surveillance state but it will meet the stated goals of those pushing these measures.
Keeping children off of social media is a much trickier problem. I think we'd be better served by banning certain sorts of algorithmic feeds.
Okay, so make it illegal for them to record any information which is what the actual laws do (or better, explicitly criminalize all the other current stalking). The point is you don't need to be prescriptive about how to prevent children from accessing the sites. Just make it so you can face massive fines and be arrested if you don't. They can figure out how to comply with the law, and they can be effective or be shut down.
They're not actually owed a solution for how to make their business model work. They can just be told that what they're doing is unacceptable, and they can figure out what they'd like to do next. If you're worried they might react with some other unacceptable thing, we can clarify that that's not okay either.