Every start-up that relies on someone else (eg. Twitter) for part of their core competency is subject to the whims of that company. The risk of that happening is something that every developer must consider. Places like the App Store are relatively safe, one would think. Google closing down services certainly brings up questions as far as developing an API utilizing less popular services.
This reduction also creates opportunities. For example, Evernote made it easy to import Google Notebook data, thus gaining off of Google axing Notebook.
Google, and developers who make use of their APIs, are in a real catch-22 there. The service can't become popular unless developers make use of it and help grow it, but you don't want to spend time developing against an API that isn't already large enough that it's not going to go away.
I always thought/hoped that Google had a large enough and mature enough infrastructure (including people, process, tech) that it could support a lot of essentially one-off services (at least as far as income generating goes) because the core services are so big. And that this infrastructure would give these smaller things time to incubate and gain significant traction. There's something to be said for a small service that's been running and been available forever, that its existence is stable and the choice is secure. Things don't seem to be getting a chance to properly incubate though.
