[Posted on behalf of Rob Wilson]
I found the Harper Collins map story a fascinating example of lack of peer-reviewed/cited data in public domain... and the story and ramifications are still unfolding. May I suggest someone more knowledgeable than me considers writing another AGU EOS article about data peer review issues using this as the prime example. It's a simple real-world issue on something as basic and publicly available as an atlas that can be related to by everyone. Plus a showing a wiggly line of data to prove your point is less inviting than a map - everyone loves maps.
Here's a current summary I found from the Guardian website: http://www.guardian.co.uk/world/2011/sep/22/times-atlas-reviews-greenland-map?newsfeed=true
I also recall there was confusion over some of my terminology last night - which was completely fair. My peer review of the PDS data set was largely an auditing thing*, checking files were in the correct format, correct number of characters and such. My format checking was described as good data stewardship or such (I forget the exact phrase) rather than what might be considered as peer-review of a data set.
However I'd say it is peer review. Using the Journal analogy again, one of the criteria requested for the reviewer to comment on is essentially 'is the English of an acceptable level'. This was summarized to me in an email of a recent AGU GRL article I reviewed as: "> Needs editing for grammar: No"
To me, getting the data to be the correct number of characters per line and of the correct format is the equivalent of grammar and writing proper English.
Still could be considered auditing, but I still think it counts as peer review - but perhaps 'peer-reviewed auditing'.