Not sure if you’re aware  of this or not but normal url shortening services like Bit.ly and tinyurl use a 301 redirect,  and the 301 redirect passes the actual  referrer of the page the bit.ly link was on and that’s why you see twitter traffic in your Google Analytics account, but I bet it  never matches with the real traffic  :

which works fine if the user clicks the link on twitter.com But more and more people don’t do that they use firefox plugins and iphone apps, which destroy the twitter referrer

Which is why you always have that gap in your stats and should see more and more direct hits, but the big issue is when you have custom url shorteners :

You can fake the referrer. Now this is great and bad for a few reasons one Google like to know the upstream and downstream, merchants like to know where your aff traffic is coming from

Imagine if you faked your adwords traffic to a merchant you could cause untold issues especially if they bid on those hard earn keywords you have found

DaveN