The RSS feed parses what it's supposed to; their feed displays data in a manner that isn't what the RSS spec says it should.
As for the ATOM feed, this script isn't designed to parse that at all... That's why it's politely ignored.
Basically it reads the title, description, and pubdate; description should hold data about the link, the title is the title of the story, and the pubdate should be the date associated with it.