What seems to be missing from this article: Mark Steyn, a conservative talk show host, called Mann a fraud. So, Mann is suing Steyn for defamation. As his defense, Steyn is trying to prove that the data was manipulated and cherry picked. Therefore, proving that Steyn's comments were justified. So, Steyn requested the data under the FIOA, since Mann's work was publicly funded.
But Mann - the scientist who warns us that global warming is real and dangerous based on a computer model - refuses to give out the computer code and data that he used to form his assertions. To me, this doesn't sound very scientific or very honest.
This highlights one of the problems with current science.
Early on (1960's) scientist used to include their code at the end of their paper, for all to evaluate. Today, predictive codes are significantly more complex and have grown into a real cash cow for the scientist, if they catch on. Thus, they are no longer distributed for free because the scientist wants to keep the code internal to their group to maintain funding.
This creates a real problem from a peer-review perspective because you can never really figure out what the code is doing or if it has been fudged. At best, you get a few half-ass validation efforts that are published in some crappy journal. After that, the authors just refer to that crappy article as proof that their code work perfectly and feel they never have to justify its accuracy again. (The editors feel that way too!)
For highly controversial and groundbreaking studies, I would argue that all the data and code needs to be clearly laid out after acceptance of the paper. The data needs to be available for people to analyze on their own. The code needs to be available for a more rigorous peer review. I understand that that may pose a hardship on the scientist for future studies as it gives away their "edge," but he has clearly reaped tremendous press over his paper so maybe it balances out.
My personal opinion (aside from what I think of global warming) is that the fact that Mann is unwilling to release his code or data supporting his famous paper probably indicates that he is at least a little bit worried that the code is wrong or some data was cherry picked. That's not really a very good stance for an academic who is making a groundbreaking argument.