Check out this slow mo vid by Dr. Dave... The first shot is a draw shot which clearly shows the tip make contact, compress (so nearly half the tip is in contact with the ball), then ride down the ball as Bob pointed out (as in remains in contact at same point of contact but releases lower, not sliding on to different parts of the ball....which is how I erroneously misread Bob's post at first).
Either way, it appears as though, once you factor in tip deformation at contact, the contact patch is much larger than the 4mm you used above.
Regarding the miscue zone, it does seem like a rounder tip would allow you to ride around the ball a bit more without that partial miscue happening compared to a nickel shaped tip. Whether that amounts to more than JJ's 5.77% I don't know.