First I want to illustrate my point by showing a Scott Addict's headtube.
As you can see, the downtube's junction on the headtube is fairly high, and the fork has a crown as tall as Yao Ming.
This creates 2 issues:
1. The front end stiffness suffers, because the two headset bearings will be closer for a given frame size, and that doesn't give you great torsional stiffness.
2. The aerodynamics suffers as there is a lot of area between the downtube and the crown junction.
My approach (and this is nothing new, Cervelo is already doing it) is to spread the bearings as far apart as possible, including depressing the fork crown. My fork length will be 355mm-ish... although I'd have to find out if manufacturing will run into any issues...
Re point 2, Pinarello/Merckx are doing something interesting to smooth out the airflow around that area.
This is a good move as we have seen wind tunnel data that indicate the high air turmoil around this junction. But their "fairing" after the fork appears to be pure dead weight and does not add to the structural soundness of the area.
So a "lower the downtube" with a modified shape will fix that. The enlarged section will also contribute to the front end stiffness. In addition, this will effectively place the lower bearing near the middle of the downtube junction rather than below it, which should have some stiffness benefit (however little).
Lastly, a lower downtube junction gives you a slightly shorter downtube as well (probably 3mm). No biggie but this is WW, so there you go...
I better catch up on 80s rap music. 
