Burton G Malkiel and Atanu Saha published a report in 2005 looking at hedge fund returns, persistence of hedge fund returns, and variance of those returns. The work was supported by Princeton's Center for Economic Policy Studies.
The "persistence of returns" figure caught my eye. Essentially, Malkiel and Saha asked "If I am fortunate enough to invest in a hedge fund that displays better than average returns for a year, what are the odds that that same hedge fund will persist with better than average returns for the following year?"
In their 2005 report, Burton G. Malkiel and Atanu Saha claimed "We found similar results for the entire 1996-2003 period. Indeed, the probability of observing repeat winners during during the period was basically 50-50."
So there you go. Hedge funds provide low correlation, but high variance in returns and a 50/50 chance of persistently above-average returns. They have a high barrier to entry, are very illiquid, are lightly regulated, and lately seem to be a lightning rod for types like Madoff.