Note his hypothesis that
condos are the core of the problem -- they aren't rented, they're purchased with credit by young professionals who "spend the creditworthiness" they gained by moving to a creative-class city, and live in them until they sell them to other incoming young professionals. There could be plenty of
rental demand that isn't affected by building lots of dense primary-residence condos that just attract more of the same. He does wonder at the end what would happen if you built so much housing that you broke that cycle, but that would be hard.
Basically he's saying there are two tiers of housing consumers, and building dense condos does nothing at all for the lower tier who can never bridge the creditworthiness gap to get into them.