There are two other reasons bike lanes don't have the kind of demand induction properties that roads do:
-Bike lanes (and ultimately, bike destinations) have way higher humans-per-square-foot of road / parking lot density than cars lanes. The throttling mechanism on behavior for the car example is ultimately drive time, and new lanes quickly become capacity constrained (first at the interchanges; later at the parking; last in the lanes themselves) in a way that slows ultimate travel back to the indifference equilibrium. The equivalent for bikes tolerates a way higher flow of humans.
-Induced demand for cars is in part a function of the fact that you can (up to a point) drive at any speed, meaning that if roads are added that support commuting in from 30, 40, 60 miles away, that can be a doable commute. There is no amount of development that will create a 60 mile bicycle commute. Here, the demand induction mechanism with travel lanes and housing is reversed: you need convenient housing to drive the demand for bike lanes.
-Bike lanes (and ultimately, bike destinations) have way higher humans-per-square-foot of road / parking lot density than cars lanes. The throttling mechanism on behavior for the car example is ultimately drive time, and new lanes quickly become capacity constrained (first at the interchanges; later at the parking; last in the lanes themselves) in a way that slows ultimate travel back to the indifference equilibrium. The equivalent for bikes tolerates a way higher flow of humans.
-Induced demand for cars is in part a function of the fact that you can (up to a point) drive at any speed, meaning that if roads are added that support commuting in from 30, 40, 60 miles away, that can be a doable commute. There is no amount of development that will create a 60 mile bicycle commute. Here, the demand induction mechanism with travel lanes and housing is reversed: you need convenient housing to drive the demand for bike lanes.