And you need to make sure there's a big talent pool there, and good universities.
A few places are close to that, but it's not so easy to replicate.
This book talks about it some: https://amzn.to/2XAU3f0
And most of the places that are close have rampant NIMBYism problems of their own.
Sure, some jobs will go elsewhere, but there are big returns to being where everyone else is. This book does a good job explaining why:
There are 'network effects' at least in terms of jobs and that environment. You'd have to convince a bunch of VC's to move to wherever along with a bunch of tech workers.
Pretty much it is a networking effect issue. The same reason why you are posting this on hackernews instead of reddit. Physical proximity matters a lot, even when work is digital.
I suppose that compared to Tokyo, most anything could be considered a small town, but in Italy, Bologna isn't.
Quibbling about details aside, I think your point is a good one, although I also believe there are definitely two sides to it. The case made in this book is convincing that cities are a lot better for the sort of "spontaneous idea contamination" that can lead to big things:
Things get even more complicated when families come into the picture: a beach town in Morocco is not my own idea of the place I'd like to live with mine, although I certainly wouldn't mind an extended vacation there.
There are a lot of things I don't care for about my hometown in Oregon (THE WEATHER!), but I do find that I'm pretty partial to the mid-sized (which for me is something like 100K-400K, depending on various factors) university town like that where I grew up. I like being able to chat with people about programming over drinks from time to time, or talk about business, or have a variety of local businesses. On the other hand, with a family and not wanting to work for a BigCo, I'm not really interested in big cities any more.
He talks about why things have not gone 'flat'.
Get dozens of book recommendations delivered straight to your inbox every Thursday.