Toronto

I think maybe he meant it was a critical mass thing. If you half-ass your cycling and pedestrian infrastructure (ie.: basically all of North America), you remove or narrow motor vehicle lanes. Few people ditch their cars for cycling or walking because the infrastructure still sucks, so all you end up doing is having a bit less space for practically the same number of cars.

Either you make the cycling infrastructure usable and comprehensive or it doesn't work at all.

The whole argument around induced demand is that people have a certain 'tolerance' when choosing a mode of transportation, and if an option becomes intolerable then they'll switch modes to a more tolerable option or forego trips entirely. If enough people mode shift to public transit and cycling then driving becomes faster and some people will switch back to driving or drive for more trips they otherwise wouldn't take until an equilibrium is once again reached between speed and tolerance, so the author's argument is the only way to ensure this equilibrium is reached at speeds that drivers desire is to reduce the number of people in the city overall.

Since this isn't a realistic option, the author argues that 'fixing traffic' needs to be re-framed as 'improving the quality of life for people.' As long as politicians are obsessed with 'fixing traffic' via increasing car speeds and reducing delays for drivers things like cycle tracks and transit lanes are a political albatross around their neck, but if politicians instead frame the objective as improving the quality of life for people moving around the city then things like building more transit lanes, cycle tracks, pedestrianized streets, etc. become more reasonable and justifiable.