My own dream as candidate/interviewer is to spend some hours trying to solve an interesting problem together that neither of us has solved before. Unfortunately it's not repeatable for the interviewer (at least once it gets solved, until then there might be a way with a big problem to bootstrap a candidate to build on past candidate+interviewer's efforts), and even then its main use is judging "works with the interviewer well" when there are probably better things you should be measuring in most cases.
It simply said "Make a working clone of this webpage, zip it up and file it to us".
I could have just nicked the jQuery code (this was before a lot of the frameworks have taken off and everyone abused jQuery and I worked at a place that used vanilla js for performance).
This is for two reasons
1. I don't like silly games in interviews. You aren't respecting me or my time before I am working for you, you aren't likely to while I am working for you.
2. I personally like to be left alone while working on something. Then again I find scrum calls etc a total waste of time.
This could actually be the test in itself. There seems to be a substantial contigent who believe that this is what a collaborative programming environment looks like, and they might actively want to select against people with a more shut-the-door-and-work-through-the-problem mindset. I'd argue that a pair programming exercise is probably better than whiteboarding-with-interruptions, but setting that up well is more work for the interviewer.
Edit: completely messed up the quoting in original!