Its hard to test a very large scale theory based on optimism about human nature, but you can downscale it and try it on smaller scales like states and cities. So why didn't your example of Instagram benefit West Virginia or Detroit?
If a self organizing phenomena doesn't happen at a small scale in almost all examples, you need a theoretical reason why it doesn't downscale if your claim is some time in the future it'll self organize at a larger scale.
The future of the whole country probably resembles WV or Detroit much more than SV. This has certain startup implications; for example, creating a luxury electric car startup in a permanently down-trending country seems unwise, in comparison sharing photos sounds like something poorer people can continue to cheaply do as they gradually become permanently poorer and poorer because even the 3rd world has mobile phones.
Exactly, anything that scales up to large, especially to fast ends up becoming an inefficient and uncontrollable behemoth. Take a look at any corporation(ie. Microsoft, HP, IBM) or governments(USG, China vs. Canada, Poland, Finland) or media (News, MTV, Cable) where large entities cannot function efficiently and only get worse as they become larger.
The only thing that works is some sort of self-organizing nature that similar to the way biological organisms form, which then end up creating neuron/brain like structures which form higher intelligence. Take a look at the universe, there is an interesting talk on how the Internet is starting to show that phenomenon, it's appearing more and more like a brain and may be starting to show some signs intelligence.
Take a look at this post from LinuxCon2013(it's not about Linux).
The conditions for that to happen aren't there yet. There's a growing population of unemployed people but it's not like there's no enough cash flow that companies are struggling to find customers. I think if and when unemployment will be so widespread as to pose a real threat to cash flow we'll see countermeasures to what are now just bleak predictions.
If a self organizing phenomena doesn't happen at a small scale in almost all examples, you need a theoretical reason why it doesn't downscale if your claim is some time in the future it'll self organize at a larger scale.
The future of the whole country probably resembles WV or Detroit much more than SV. This has certain startup implications; for example, creating a luxury electric car startup in a permanently down-trending country seems unwise, in comparison sharing photos sounds like something poorer people can continue to cheaply do as they gradually become permanently poorer and poorer because even the 3rd world has mobile phones.