As a defender of good data makes good conclusions, I find myself in the uncomfortable spot of agreeing with an opinion piece put out by the National Association of Realtors (NAR). This opinion piece starts to get at the real problem with Case-Shiller's data. I would argue that the Case-Shiller indexes have far too high of a bias towards subprime loans which heavily skews the data.
They link to a paper created by OFHEO that tries to reconcile their results to Case-Shiller's. I'll report on any findings after I've had time to digest.