If they could handle it using minimal automation, they would. They can't, so they don't.
Figuring out which real human owns an account that was never verified to belong to that human in the first place is not easy to automate, if possible at all. It's expensive, error-prone, and time-consuming. And if not done carefully, it makes it far easier to steal accounts than it already is.
I understand that many commenters have magic wands. Unfortunately, Meta does not.
> That they want to be sued?
Q.E.D. It has been demonstrated by their system design.