Those Atlantic articles seem like bogus trend pieces.

Certainly. I think there's multiple overlapping trends feeding the final results, but the cited stats could be purely cyclical or cherrypicked.

Suburban populations are still rising, just relatively slowly compared to the growth of cities.

And the total population is still growing, which means a shift.

And driverless cars are a boon especially to suburban areas, since they make commuting less annoying (and potentially, much faster).

This can only happen after autonomous cars are accepted and widespread for ordinary driving, hence it doesn't matter to my argument about acceptance.