That is a pretty nice response, even if it is a bit snarky :-)

I don't think you, Daniel, would do well there. When I was interviewing people at Google I would always try to remind the candidate that whether or not the interview proceeded to the offer stage said more about the people who interviewed them than it did about the candidate's capabilities.

I like the refer.ly question because it allows the candidate to talk about something they are passionate about, or have thought deeply about. I'm not sure that it informs on their organization as much as the interviewer would like.

Be aware that you incorrectly interpret results of the question: all candidates guess correctly, but only half of them do it in your Universe.

Well, let's be fair... the method the article describes is a bit better than random. It's just asking for a work sample, except the work you are asking for is efficient communication.

I doubt the "inventor" of this method really thought it was the one true question, we can attribute that tone to the author of the article.

It's refreshing to see that some people have an ounce of sense when it comes to interviewing :)