The sensible regulation would be aiming at socially harmful features of tech, not forcing people to provide proof of identity to a third party in order to continue harming themselves with the same exact features.

Imagine if such an approach was taken to, for example, food safety? Instead of closing down a restaurant that has poor hygiene, you'd be instead horce the restaurant to hire a private security contractor to check people's IDs to verify that they are old enough to consent to getting a foodborne illness. That's an absurd approach.