In other words, just because urban designs correlate with travel behavior, it doesn't mean they cause it.
This view is partly true.
Urban design does cause massive changes in travel behaviour, it's just that urban design isn't the only factor causing change in behaviour. Money and time are the two biggest other factors. You try making that highway more expensive (toll booth) or reduce its efficiency, and people will immediately search for different routes to work. But getting people to change modality (mode of transport) takes an astronomical effort.
People are creatures of habit. Good luck getting those 300-steps-a-day Americans out of the car. Not only does it take a reliable, faster, more comfortable mode of transport (e.g. the Lightrail hype currently), you also need to change their daily rhythm. It is far easier to create an excuse for inaction than a convincing case for action. A train could save time, effort, environmental footprint and still people will dismiss it entirely 'because trains are for poor people' or whatever small excuse they come up with.
The article addresses one of the big questions in planning: how does the built environment change behaviour, and more interestingly, how do we make changes in the design of cities to change behaviour for the better? What is 'better', for whom? These questions don't have straightforward answers, as there are just too many variables to take into account. But single-zone land use has had its day. Large commercial, industrial and residential districts only forces people to travel by car. Combine it with low density and you have the deadly cocktail that degrades not just health as the title suggests but also relationships, time, etcetera.
You only need to travel to get something that isn't where you are. So if you have your daily needs (work, school, shops) around the corner, like in the Atlantic Station example, you won't need your car. And since there are no pre-existing habits when you move somewhere else, it's easy to forget about the car altogether.