We all have too much to read, so here's a tip to save time. Whenever any article (such as this one) cites information about incorporated US cities as a basis for any claim about trends in the culture, quit reading. US big-city boundaries are irrelevant to most people's lives, and to anything else that matters about our culture, economy, or destiny.
Christopher Leinberger makes this point in a New Republic article recently, usefully expanded on by Sarah Goodyear at Grist. Leinberger argues that "city" and "suburb" is no longer a useful opposition, and that what really matters are walkable urban places vs drivable suburban ones. True enough, but replacing city with it's near-synonym urban doesn't take us far. "City" and "suburb" are rich, evocative, and succinct words. The word city in particular must be fought for, redefined in ways that defend its profound cultural heritage. The word has an ancient and clear lineage from Latin, one that forms the basis for the word citizen, not to mention civic and civilization.
Greek and Roman political theory was all about the city, in a sense of that word that we can recognize today: groups of people living together in a small space for reasons of security and economy, but also the site of humanity's cultural and intellectual development. City is a word of enormous evocative power to capture a range of ideas that drive urbanism. Leinberger himself can't describe what really matters without using the word urban, which evokes a similar history and resonance.
What Leinberger is really complaining about are discussions of data about incorporated US cities, which are a very narrow and specific problem. A few of the oldest US cities (San Francisco, St. Louis) have coherent boundaries that describe real cultural and demographic units, but many are bizarre shapes of purely historical interest.
Nobody who understands the lived experience of Los Angeles would claim that the City of Los Angeles is a useful or interesting demographic unit. While the city excludes a great deal of dense inner-city fabric close to downtown, it has long balloonlike tentacles extending north to take in the whole San Fernando Valley and also south to grab the port of San Pedro. It also contains a good deal of near-wilderness in the Santa Monica Mountains.
The tentacular, pockmarked, pulsating blob that we call the City of Los Angeles is the map of a long-ago war over water and power. The only people who care about it today are those who work for city government or serve as its elected officials, plus a few who've considered city taxes and services as a reason to locate in the city or out of it.
Americans should notice, too, that bizarre and misleading city boundaries are largely a US phenomenon. Europe, Australia, and New Zealand generally allow central (state or national) governments to draw the boundaries of their local governments, so these boundaries usually (not always) end up making some kind of sense. (With the exception of Queensland, Australian local government areas are too small to have much influence, but that's a different problem.)
As Leinberger says, we need a distinction between walkable urban communities and drivable suburban ones, and American city limits are useless for understanding that distinction. But the word city -- whose Latin ancestor meant "walkable urban" for millennia until about 1950 -- is still worth fighting for. Legal US "city limits" are an imperfect and aspirational approximation of what cities really are, and what they really mean for the human project. Despite their pedantic misuse by the likes of Cox and Kotkin, city limits have no authority to tell us what a city is, and why we should want to live in a real city or not. The deep attractions and repulsions that we feel for big cities are the key to a longer and truer cultural understanding of what cities are, and of why the civic is the root of civilization.