I think "old cities" are pretty good too: the kind where they were built for people, rather than cars, and where you could build things without a dense thicket of regulations almost entirely unrelated to safety. Another thing that seems to work well is limiting the amount of up-front large scale projects and growing incrementally. No one can plan for everything ( https://en.wikipedia.org/wiki/Economic_calculation_pro…
Cars have a place in cities but maybe they should be in tunnels underground or something. I view money locked up in real estate as money in a very suboptimal place. We pay too much 'rent' in general; it'd be far better to try to make housing/office space cheap and have people invest the extra money in productive assets.
In the U.S., we've collectively done a lot, culturally and legally, to make real estate a principal savings vehicle for many if not most people, and this has had a lot of negative consequences. Zoning Rules!: The Economics of Land Use Regulation (http://www.amazon.com/Zoning-Rules-Economics-Land-Regulation...) by is a good introduction to some of them. Chapters 7 – 9 in particular explain how the 1970s saw the growth of zoning restrictions that have driven housing costs in urban areas relentlessly up.
I'd be curious about alternate land-ownership structures or incentives that make people view housing more as a service than as a good store of value. I don't know what that might mean in practice, but the idea itself seems important to me.