Matt Yglesias' claim that the difference between Pittsburgh and Detroit one has major universities downtown and the other does not sounds plausible, since it describes one dimension on which the two cities are different. (It neglects other relevant phenomena, like Pittsburgh being in Pennsylvania and Detroit being in Michigan, which might explain the underlying dynamic better, but never mind). But the claim itself is unsustainable. To wit:
The argument is false because high-level universities do not save even their neighborhoods: the University of Chicago can barely (and with marginal success) gentrify Hyde Park, and has close to no impact on the adjacent neighborhoods. USC is in an unsafe part of LA. The status of each school, its medical campus, research facilities, etc, is unable to sustain even a small urban environment.
The argument is false because it gets the causality backward: New York City has Columbia, NYU, and a host of other smaller colleges and universities, but no one would be so foolish as to argue that the universities themselves played some role in drawing people to New York. Indeed, one could plausibly argue the causality runs the other way: the universities are successful because the city is successful.
The argument is false because the growth and size of cities has nothing to do with the universities it contains. The top ten cities by population: New York, LA, Chicago, Houston, Philadelphia, Phoenix, San Antonio, San Diego, Dallas, San Jose. How many of these depend, in substantial part, on the presence of a university?
The argument is false because it can't even be bothered to compare two comparable things: Detroit, even now, has over twice as many citizens as Pittsburgh. Pittsburgh requires two major research universities; by Yglesias' logic, this would require Detroit to have four--not just moving the University of Michigan to Detroit for some reason.
Which leads us to the thing the argument might prove: is mid-sized cities--say 100,000-300,000 or so--the presence of a few major research universities might be enough to function as an economic driver for the area. That would explain Ann Arbor's ability to resist Michigan's economic downturn, and the ability of the Triangle to keep steady in North Carolina--three major research universities support an industrial base.