All of which is a throat-clearing way of saying that if we see a big increase in the amount of walkable urbanism available to American families, an awful lot of it will probably exist outside the city limits of the municipalities that form the hubs of our metropolitan area. That will mean, yes, converting existing elements of the build environment rather than simply abandoning everything and trying to get everyone to move willy-nilly into downtown Cleveland. In other words — more housing in malls.
But the real kicker, in my opinion, is going to be getting useful jobs into these areas. If all you have is a bunch of rich people who live at the mall and shop there while working 20 or 30 miles away by non-mass transit, and a bunch of lower-middle types who work at the mall and its associated offices while living 20 or 30 miles the other way, also by non-mass transit (because you wouldn’t want to make it easy for “those people” to get to the upscale mall and its residences), then you haven’t gotten very far at all.
The Tysons area may have done this pretty well, with oodles of office towers in bowshot of the malls and the incredibly ugly apartment blocks, except for the problem of getting from one part of the place to any other. You get in your car, drive out of the parking lot, navigate a couple of eight-lane cloverleafs (might only take half an hour if you’re lucky), find another parking spot and walk to your destination. If they had shuttles running continuously, or a few monorail loops, or just put up enough pedestrian bridges to pave over the big honking roads, it would be just perfect.
In the comments to Matt’s piece there’s a little bit along this line, with someone noting that this same kind of closely-sited residential and retail develpment at Springfield Mall/Kingstowne, but with a few hundred yards of completely unwalkable distance between them. Pedestrian bridges and their equivalents really aren’t so hard to build, but many suburban developments seem religiously opposed to them.