> 1. You fail or have meager success;
I don't know why people keep repeating this fallacy. There is no shortage of counter-examples, the most obvious being Zynga (hell, the homepage has a story about Angry Birds' massive success _right now_, which is another example).
And if you don't think Zynga is at risk from a change of policy by Facebook you're kidding yourself.
It's a gamble. A huge gamble. Zynga may in fact be large enough now to survive war with Facebook but they'd be savaged and greatly diminished by it (IMHO).
Look no further than the move to Facebook Credits [1] to see who really wields the power in this relationship. Facebook has exacted a 30% cut, forced on app developers (interestingly I haven't seen quite the same outrage at this as I have about iOS subscription changes).
[1]: http://www.allfacebook.com/zynga-facebook-credit-2010-09
I would say a great deal of the companies in the world are somehow at the mercy of the fates, in a way similar to Zynga. A bus stop moves, and your convenience store no longer gets traffic for example. Lockheed Martin has one major customer - what would happen if the DoD arbitrarily changed the rules for them? Every tech company relies on Google and Twitter for marketing and traffic, what if they got caught it some algorithmic change and stopped getting search traffic?
And so what - there's risk everywhere, but there's opportunity too. At the risk of getting screwed, Zynga carved a $200m a year business. And now what? They get 30% less? A nice problem to have.
No, it's not. Ever.
Twitter don't think of third-party apps as any benefit to them, whereas Zynga was a boon to Facebook.
I think it does come down to how big of an asshole the ecosystem provider is. Microsoft, Apple, Facebook, Twitter, etc have all done this. And they have to weigh fucking over a partner against the benefit to their core product from muscling in on their partner's turf.
As a developer, people have been asking for years "How valuable would this be as a part of Windows?" and "How big a dick is Microsoft?" and "Do we have any competitive advantages over Microsoft?" and deciding whether it's worth taking the risk to build a particular business on Windows.
But in the end, you've got to build on someone else's platform, whether it's Windows or Twitter or iOS... or as in Mahalo's case, depending on Google for your traffic.
If your business is built entirely on someone else's ecosystem there are two possible outcomes:
1. You fail or have meager success;
2. You're very successful at which point the ecosystem provider will probably judge what you do important enough to acquire you (or one of your competitors!) or compete with you and have the home advantage.
The fact that some convinced themselves that Twitter was "different" might want to refer to the Peanuts theme of Lucy and the football.