Lots of pages on The Onion display content fine in the Google Cache, but if you visit the page itself – even when you're coming from a Google result – you get a "not authorized" message. E.g.
http://www.theonion.com/content/node/30649

Is this cloaking, or did these pages simply change their content recently, and the Googlebot hasn't yet updated its crawl of them?

Looks like something maybe have just been updated which has probably cocked things up. Check this page:

http://www.theonion.com/content/node/30656

There's a link to the "CIA Asks Bush To Discontinue Blog" story at the bottom which returns the same page you linked to with the "not authorized" message.

Incidentally, the print version of the same article works fine:

http://www.theonion.com/content/node/30649/print/

... and that just made me realise something!

http://www.theonion.com/content/node/30649 – broken
vs.
http://www.theonion.com/content/node/30649/ – ok

When you don't add the trailing slash to the URL, it breaks.

What's even more odd is that it's completely the opposite for this URL:

http://www.theonion.com/content/node/30656 – ok
vs.
http://www.theonion.com/content/node/30656/ – broken

And it appears to be fixed now already!

Heh. Now that's quickly solved :)