Hacker News Reader

In part because scraping the website is going to break all the time, and sometimes on things you simply can't see because they're running an a/b test. Then there's usually the site TOS that explicitly denies scraping (I don't know if Twitter does, but it's standard fare).

Twitter want vertical integration.

They're not quite sure what they can do with it, but they know any hope of monetization is lost if all they are doing is providing a backend for third party clients.

That's a very nice sentiment, but it won't fly if you're trying to provide third party authentication in a manner that minimizes the potential for fraud (hence, the oauth clusterfuck).