"About 90% of the time, the [waterfall] process results in a late, over-budget, fragile, and expensive-to maintain software system. A typical result of following the waterfall model is that integration and testing consume too much time and effort relative to the other software development activities. Most waterfall projects, expend over 40% of their effort and schedule in integration and testing."

Oh well, maybe we should stick with what we know after all :)

(Thanks Grady Booch and David Ing blogs for the conference link)