I think it's entirely reasonable that they have at least one review department per market that they want to serve. Again, that's the way the ad business worked until very, very recently. It's not like some onerous idea made up just for them.

You're drawing a false dichotomy about censorship. There's a normal, accepted middle ground between oppressive censorship and anarchy, often referred to as "ad checking" or "editorial review". Right now I can use an automated system to buy an ad saying "Trump Eats Babies" or "Obama Isn't a Citizen" or other flat-out lies. This is wrong. There should be a human in the mix who does at least minimal verification.