Eh, I think they hit the nail on the head in the article. Having no clear direction and too many changes is a very easy way to kill the development of any project, much less anything that large. It’s kind of a reoccurring theme in long-running, canceled or badly implemented projects these days (Bioware comes to mind).