It's just Jevon's paradox. High prices suppress demand because possible use cases are infeasible/unprofitable, but lower costs make them possible until you have total oversaturation (but e.g. subsidized US corn results in it used in all products!)
Indeed, not only is total demand a function of cost, but so is frequency of that demand. If my analogy were an actual hypothesis, it could be tested by checking that those who live closer to a grocery store will have on average a lower grocery bill (per visit) than those that live farther away (normalized by the number of people per household and household income), and the absolute total number of trips taken by those that live close by will be higher.
Which is to say, it's profitable to go to a grocery store for very few items if you live close by. Live farther away, and it's less profitable, perhaps to the point where it's not worth doing unless you buy a lot.
I agree that latent demand becomes a factor for sure as well.
