"Urban sprawl" is vague and used to score rhetorical points. Auto-oriented development is much clearer. So now the NY Times introduces "vertical sprawl" in "Cities Grow Up, And Some See Sprawl."
The trouble with the new discussion is that no one has a clue as to which settlement densities are either "too high" or "too low" at a multitude of sites -- for a multitude of time periods. There is no science here. This is why we have land markets.
Strangely, however, the idea of land markets is never acknowledged and industrial policy is presumed. This introduces politics and conflict and rhetoric. It provides fodder for oceans of (seemingly) learned discussion.
Progress would involve moving away from conflict and politics by expanding the opportunities for exchange. Even externalities management is now widely done via tradeable and marketable rights.
Markets are exotic (and vaguely sinister) in the mindset of the best and brightest. And it is much worse when it comes to land markets.