MajorThumb

I'm in the UK and my order arrived today with no dispatch email.

First, I know someone else pointed this out, but the "pay what you want" experiment was always intended to be a one and done type of deal.  The band made that clear very early on.  It wasn't intended to be repeated so it doesn't really make sense to try and equate "success" with them trying the same thing in the future.
Second that's entirely speculative and depends on your personal stance on what success means.  You can't really speak for the band so you can't gauge exactly how well it worked out for them.  Everything I've read from the band indicates they've been more than happy with the results and as the other link posted mentioned they were more successful digitally than they were with their previous releases combined.  It sure sounds like a success to me.
 
 
 
See that's the thing about this "debate", no one knows what the average price paid actually was because the band hasn't given any details.  According to the only "data" that we do have on the release, the "average price paid" for the download was indeed $6.  If you have another source for your estimations I'd be curious to read it.  I think it comes down to a misunderstanding on the wording.
The closest to "data" we have is an internet monitoring company named Comscore (mentioned in that first link).  According to them the average price paid was $6 ($8.05 in the US).  BUT, and perhaps this is where the confusion might be, the average price per download was $2.26.  This is because Comscore believes 62% of users did not pay anything.  So when you tally up the 38% who did pay the average price paid is $6, when you add in the people who downloaded for free the price drops to $2.26.  To put it simply, if you paid nothing you aren't part of the "average price paid"... you didn't pay.
http://www.comscore.com/Insights/Press-Releases/2007/11/Radiohead-Downloads
At the end of the day I see a release that garnered them more profit than their previous releases combined digitally, a physical release that totaled 1.75 million in the first year alone and a 100,000 disc box sets sold at $80.  Plus, they actually got to keep ownership of their own material.  If that isn't a success, I really don't know what is.  That's my opinion.  And that's what I think is the coolest thing about the experiment.  It doesn't really matter what we think, what the label thinks or what Billboard thinks... they managed to avoid all that.  In the end they remain the only ones who really know the details on it all, and fittingly for one of their first times as a band they would be able to decide what success really means for themselves.