So, I believe that startup cities are those that amass a significant amount of potential to facilitate collision density between different kinds of thinkers. Collision density is naturally much higher in cities than in suburbs and suburban tech parks because cities house a much higher rate of diverse actors in smaller, more compact spaces. As the tools of innovation are becoming more democratized, and cities are increasingly home to dozens, if not hundreds, of accelerators, coworking spaces, fab labs, and the like, combined with proximity to artists, designers, schools, and users, entrepreneurs are able to regularly mingle with and create mashups of ideas, some of which may turn into commercializable innovations.
Emphasis added. The words "believe" and "naturally" act in tandem in this non-argument. Naturally, everyone believes collision density is higher in cities than in suburbs because Jane Jacobs didn't like sprawl. That's how the paragraph reads to me.
For geographers, the use of the term "natural" or "naturally" will set off the alarm of environmental determinism. Many geographers believe that environmental determinism is bad. The perspective is an academic taboo, like geopolitics. Naturally, Nazi Germany should do something about that Czechoslovakian knife sticking in the side of the Vaterland and keep going until lebensraum is restored.
The United States has a long, recent history of successful innovation and commercialization in the very places where, naturally, collision density is lacking. Believing in Jane Jacobs density won't change that fact. But democratization will? Excuse me while I pursue my manifest destiny.