Hold on. The closest thing to an empirical test that anyone has ever done, and published. Carbon was in its infancy. The people doing the test were expecting steel to do well. And steel completely shat the bed. As in not even close. You can feel free to dismiss it. I tend to think that it shows that steel is not the ideal material for lightweight, racing bikes. But hey, until someone else does another empiricalish test, (maybe in another 15 years) it's the best we got...