Facebook C# development: then and now

I have overseen my fair share of web applications over the years. And I have increasingly dealt with API's and SDK's. Two years ago, when a project required Facebook integration, I lept at the idea. It was a fairly new platform at the time, and the thought of integrating the social element into traditional web development was irresistible. One month development time and three SDK's later, the site was up and running. Having figured out all the kinks and details and gotten a working example running, you become an expert at on the subject. My thread on Github was gaining popularity as I was easily identifying where others were getting stuck. Fast forward to end of 2011 and I found myself at another Facebook web application implementation. I dreaded revisiting the old code. Combining three SDK's for one application is far from elegant. As luck would have it, much progress has been made on a comprehensive Facebook C# SDK that did the job successfully on its own. The new application is much more stable and faster, and if I ever revisit it, I don't think I'll need much to understand it. Looking at the two instances, I can't help but wonder why Facebook still chooses to operate as a startup in such instances as their SDK's? Facebook has evolved to the point of transcending being a website, it has become a platform. I think or is about time it acted as such with all the responsibilities that come with it. I understand that one of the pillars of Facebook is crowd-sourcing, but I think the connecting avenues need more centralization and cohesiveness. I wonder in another two years would we still be talking about "integrating Facebook...", or will it be pre-baked into development platforms, or will we seize to require integration altogether?