I talk a lot about height and floor area ratio restrictions as a key modality restricting the creation of high-density neighborhoods, but people sometimes point out that some very dense cities don't have much in the way of very tall buildings. Paris often comes up in this regard, but a better example for the American context is probably Somerville, Massachusetts where I happen to be at this very moment.
Somerville, a highly urbanized "suburb" of Boston contains over 18,000 people per square mile making it somewhat denser than San Francisco. And yet it has few tall buildings and certainly no skyscrapers. The key to that happening is that it's very literally dense. The streets are narrow, there's very little parkland, there are few office buildings or hotels. And most of all, the dwellings themselves are small. It's no coincidence that this town is in the oldest-settled part of the United States. Over the years as America has gotten wealthier people have tended to live in larger and larger spaces, but Somerville is full old structures that fit two or three households into apartments that cost more than the average American dwelling but have less square footage than the average American dwelling.
And as far as it goes, it's fine. But if you're thinking about a community that's not already like Somerville, it'd be somewhat curious to make this kind of a thing a model. A population density of 18,000 per square foot should be perfectly compatible with spacious dwellings, ample parks, sidewalk cafes, at least some broad boulevards, and an office district. But it couldn't all consist of two and three storey buildings. In practice, the most likely scenario is to have some very tall buildings. People put a value on building services (doormen and such) that are easier to provide if you have a lot of households in a single structure, and people also put a value on the privacy and outdoor space of a single-family home. So given 21st century construction technology and elevators, a mix of big buildings and single-family buildings makes more sense than a crowd of three-deckers.
Meanwhile, I wonder what the future holds for Somerville. This building features small dwellings in its current configuration as a two-unit rental property but as prices continue to rise in Cambridge, I wouldn't be surprised to see it bought by a rich guy and turned into a spacious single-family home. With upzoning, high demand for Somerville living would lead to new construction and higher density even as average dwelling size rises. But if larger structures can't be built, then higher demand and higher incomes are going to lead to falling density as multifamily structures are converted to single-family ones. Indeed, between 2000 and 2010 Somerville's population fell by 2.2 percent even as it's become nicer and more expensive than it was when I was in college.