I should point out that the Urbanophile post was about whether density is necessary for a city to be competitive. Briefly, businesses depend on access, and they will go where access is cheap. I honestly don't give a shit about whether my city is competitive, because that just means that someone else's city is losing out. My priorities are up at the top of the page: pollution, efficiency, carnage, society, equity - and competitiveness works against equity. (I'm aware of the value of healthy competition, but it's got limits, that's all.) But in essence, he's arguing the following:
In physics, you can increase mass while holding volume constant by increasing density, or you can increase volume while holding mass constant by decreasing density. In urbanism you can increase access while holding mobility constant by increasing density, or you can increase mobility while holding access constant by decreasing density. So Columbus might be less dense than Indianapolis, but there's just as much access because the roads are less congested. Similarly for Houston and New York, to quote a commenter.
Well yay, equitable access! But uh - pollution, efficiency and carnage? Not so great. By relying more on driving, and on longer distance trips, less dense cities promote pollution, waste money and resources, and put more people at risk of being killed or injured by cars. If all you care about is your city's competitiveness, well, how competitive can you be if your air quality sucks, you're spending a huge chunk of your budget fixing roads, and your residents are being killed?
Yonah rightly observes that this equitable access may be great for businesspeople with private cars, but it leaves out "everyone else - the young, the old, the poor, the sick," so it's not very equitable after all. Those who can't use cars need walkable neighborhoods with transit. But if people with cars can do just fine without walkability and transit, they won't support them, and everybody else will suffer. Density acts as an equalizer here, forcing walkability and transit patronage, and ensuring that walkable stores and transit providers will have enough business to stay afloat.
That's where Yonah goes off course, following the illustrious trail of Simpson, Curtin and Olsen. Like them, he treats car use as a force of nature:
Unlike inner-city districts with their medium and high-rise buildings, streetcar suburbs are characterized by low densities, little neighborhood retail within walking distance, and very few accessible jobs, three significant factors that make them difficult to adapt to transit. In other words, while they may have been built with streetcars in mind, they transitioned to the automobile age naturally.
The fact of the matter is that the absolute access afforded by streetcar suburbs is irrelevant. As Michael Kemp wrote back in 1973, when we're talking about people switching from one mode to another, what matters is their relative access, and that means comparing the access afforded by transit to that afforded by cars.
Surprisingly, Yonah talks about "failure" without discussing what failure actually means for an enterprise: people get what they want from the competition. There is no mention of how the roads got to be such stiff competition for the streetcars. It's left to Stephen Smith of Market Urbanism in the first comment to point out "the widespread subsidization of the roads, which were at that time not mostly funded by user [f]ees, but rather out of general revenues (which, perhaps not coincidentally, were in large part paid by 'traction magnates')."
People in streetcar suburbs didn't abandon the streetcars because the density was too low. They switched to driving because they could get where they wanted quicker by road, and they could do that because the government built a bunch of big new roads for them to drive on. If there had been no cars, people in those suburbs would still be taking streetcars - or they would have moved back to the cities, if the streetcar suburbs were really as inefficient as Yonah claims.
It required a massive government intervention to build all those roads, and that intervention is proving unsustainable. The roads are crumbling, the bridges are falling down, and there's no money in the budget to repair them. Despite the hugely expensive oil wars, the price of oil rose a few years ago and popped the housing bubble. Everyone who knows what's going on expects the price of oil to rise again and stay high. Eventually it will become difficult for all but the richest to own cars, and that's the main reason we should live in walkable neighborhoods with transit.
But as it becomes more expensive to own a car, transit and walking become more attractive. Transit providers and local shops will flourish, and they will return to the streetcar suburbs. In other words, the viability of streetcar suburbs is inversely related to the expense of driving.
Above, I argued that even though you could function pretty well driving around Columbus, it was not so good for pollution, efficiency and carnage. You could make a similar counterargument to my own argument about streetcar suburbs. Walking to the streetcar and the local grocery may give you good access, and it doesn't generate a significant amount of pollution or carnage, but it's not that efficient. Much more efficient to have people living above the grocery store, etc. But won't that kind of efficiency sort itself out?