Building skyscrapers is expensive. As long as both land and transport are cheap, it’s easier to just sprawl. If we come up with replicating constructors of some sort (which might just be genetically engineered bacteria and plants, as there are some thermodynamic concerns over whether strong nanotech is actually feasible) that can build skyscrapers as long as you provide them with appropriate energy and feedstocks, the value proposition for land use changes drastically.