You might consider adding the above sections, possibly on the front page where you showcase the papers using Brian in interesting ways or lesser known features with some nice examples. You could even automate this with a script to randomly draw them out of the set of papers you already list or sections of documentation (respectively). These choices could be overridden when you feel like picking examples but would prevent the website looking stale if you forget or are too busy at the time. These articless could then be tweeted to draw attention to some cool use cases or interesting models which the twitterati may have otherwise missed!