Nothing annoys me more than perfectly capable people building useless crap that nobody needs, justifying it as a learning experiment - when there are desperate and dire needs out in the world, and the people running "the official solutions" are even less qualified than you and me, all in existing markets with problems that can easily be solved, had they only been more visible.
But people don't easily share this market insight, because they think they will keep it for themselves and make money off it. On the other side, people get lots of ideas that they too think they can make money off, but that are stupid ideas and they only find out after they put too much effort into it - if they just told more people about it sooner, they would've been able to fine tune their ideas much sooner.
So I don't buy this nonsense about "creating a market". It's all about recognizing the latent market that is there already - nothing in the world is new. What is important is taking the time to look and dig real deep into the problem you think you see, before you even attempt to formulate a solution. Try to disprove your most basic assumption as early and cheaply as possible.
About Facebook - It's not a good example because it had a lot of things going for it, rich kid, rich school, rich family, good connections - the only relevant thing, perhaps, is that there were many projects like Facebook at the time, so besides connections and money, what set Facebook apart? They shared their API and allowed developers to use their data - something none of the other platforms did back then. In effect, they crowd sourced new ideas from developers the world over, all they did was to create a "standard" around which it was profitable for developers to do so. That's the revolutionary "new" thing they proved worked to gain support, but it's something that people who understand software freedom has known for years. Google did almost the same to become popular. But both of them seem to have forgotten their roots and now have shareholders who likely don't understand any of this milking and strangling them... although Mark Zuckerberg has a trump card in his controlling stake in Facebook... I wonder what he will do with it.
Finding the right environment - completely agree, if you have a real solution and it works and you find the right environment where it can take root, it will start growing... but you need to be the passionate flame burning, heating up that insanely big iron with just your little flame through consistent application.... and "Knowing when to back off" - yes, you should never become a bottleneck to the growth. Find someone who has done it before, who is keen and able, and let them. Usually people who start things are bad at running them, and vice versa... ultimately you need a good, balanced team.