Please provide empirical data.
Just to be clear, I will not accept Jakob Nielsen's Alertbox, March 19, 2000: Why You Only Need to Test with 5 Users without solid empirical data to back it up. I find it hard to believe that in 10 years no one has ever tested Nielsen's claim in a meaningful way.
Further, I find statements from Nielsen like this troubling at best: "The curve clearly shows that you need to test with at least 15 users to discover all the usability problems in the design."
Translation: Test 15 people and you will find 100% of the usability problems. (yeah, right!)