As well there's the environmental impact. Concrete creates a lot of CO2 and we actually need to use less of it going forward.
Need to get somewhere in the suburbs? Carry large items? Go in the nature for the weekend? Any route that's not been anticipated by the central-public-transport-planning-committee is probably way better done with a car latency-wise.
Above a certain income threshold, it'll be really hard to convince people to ditch their car completely. You might get away with reducing car footprint from one per adult to one per household. Any dense development should have some sort of underground parking planned for it. Then it makes it much easier to get rid of on-street parking and convert the space into safe cycling infrastructure.
For example creating a bus route to the base of the mountain so that hikers and snowshoers can access the wilderness is a big win for the public at large, but nonetheless there will always be a group of enthusiasts that want go on a further flung mountain hike, and they'll need to use a car to get there. Doesn't mean it wasn't a good idea to create the infrastructure that accommodates the bigger group.
Building big expensive concrete car lots, assuming everyone has a car, is an example of building for the edge case. In Vancouver, a city that has done not a bad job of building protected bike lanes, there's data out now to support that the parking lots that it has mandated in new apartment buildings are in fact overbuilt, under utilized and part vacant!
