Tim Bray offers a good high level explanation of why hash-bang links are a horrible idea and fundamentally break the web. I like this question:
So, my question is: what is the next great Web app that nobody’s built yet that depends on the simple link-identifies-a-resource paradigm, but that now we won’t be seeing?
That probably explains the Twitter case, since Twitter is an application that rightly has many dynamic elements. But that doesn’t make sense for Gawker, a Web publisher in the business of publishing static blog posts on the Web. Why are they loading that content dynamically? My best guess there is that they hired a developer or manager who had done it that way somewhere else, probably for more sensible reasons. They came to Gawker and decided to just build things in the way that they already understood. That person should probably be fired.