I had no idea that the failure of this plan was yet another gift from Mike Harris.
One of the big problems with most of our New Urbanist efforts is that they aren't planned around transit at all. In Europe, as a new neighbourhood is built, a transit line is built simultaneously. Cornell would have had a subway or regional rail route built right into the heart of the neighbourhood, and the neighbourhoods' central nodes would be located at the stations. That's why I have high hopes for Vaughan Centre (though I'm wary of Vaughan's past failures) and hope that it's a model that will be followed in the future. Extending regional rail (GO) into the heart of Queensville and the Bond Head development is an absolute must before the neighbourhood is actually built. It doesn't cost much to build rails on farmland, but it costs a fortune to build it through a developed area. That's why this notion that's so popular here of "progression through modes" is wrongheaded. You put the hourly bus service in at the outset, and then years later when the neighbourhood is developed try to add higher-order transit by wedging light rail rights of way down existing streets or tunnelling. It costs vastly less to put the rail line or busway in at the outset. That's how we developed neighbourhoods in the 19th and early-20th centuries, which is what New Urbanism is supposed to be replicating.
I have to say that part of the problem with Cornell retail is likely discomfort among retailers with something different. Not enough of them are willing to take a chance on locating in an urban-style development that there is no critical mass. The developers should really have pushed harder at the outset to attract a successful slate of retail.