Lancet is about to publish a study on the number of deaths in Iraq, attempting to compare the number of deaths before and after Saddam Hussein's overthrow. The study claims that there have been 650,000 "excess deaths" in the post-war period compared to business-as-usual under Saddam Hussein. Search Google News for "Lancet" and you will find plenty of information.
The study is under intense scrutiny. Asymmetrical Information has done an excellent enumeration of several legitimate criticisms and illegitimate criticisms of the study that have so far emerged so far. I will not repeat them here; go read for yourself.
I would really like to see the authors release their raw survey data. It sounds like the weakest part of the paper's argument is the data collection itself. For example, what if the interviewers simply faked a lot of the data? If the data is made public, then it can be statistically tested to see if it was made up.
Anyway, if nothing else, the study opens tantalizing possibilities for studies of the future. I would love to see the study repeated using the improvements that have come to light, e.g. comparison with official figures, recording demographic information, and releasing the raw data immediately.