The bottom line on that is test data. I spend more time testing and refining from test data than actually conceptualizing and building. I don't have any quantified data on this one yet (I need a clear day and a new digital recorder for that) but I am confident that there is a significant aerodynamic improvement. It's at least one gear better for a given perceived effort.

There are no 90 degree angle surfaces - the front is elliptical and then a slight convex shape on the sides. If you could slice a plane from the handlebars to the tail light it would look a slightly fat airfoil, about 6 to 1 length to width ratio. It is true that covering the gaps improves performance. That's a difficult compromise. At one extreme, I can't imagine trying to actually ride around in one of the fully enclosed low profile HPV. But the gaps do not eliminate aerodynamic gains; look at how street racer motorcycles are designed for example with an eye to aerodynamics. Turbulent air in the gaps can pull the air flow in so that the boundary reattaches to the surface further to the rear. It depends more on shape and angles than on the visual impact.

My decision on the compromise is to retain all of the utility and handling characteristics of a road bike as much as possible, and improve aerodynamics within those parameters. If coast-down tests don't show appreciable performance improvements I'll shift the compromise back towards the more full style.