[Patrick]

Here's what Hucho (1987, pp. 133, 135, 136) has to say about windshield angle:

As the windshield becomes flatter, the aerodynamic drag decreases. ... From all these data it can be concluded that the direct influence of windshield inclination on drag is only moderate. The effect is assumed to be more pronounced the more the flow is routed over the windshield.

Windshield inclination angles of more than 60 degrees are not practical because of light diffusion. In addition, large, highly inclined windshields lead to increased solar heating of the passenger compartment.

Two effects are responsible for the favourable, though moderate, influence of a highly inclined windshield on drag. First, the excessive speed in the area of the A-pillar is reduced so that the momentum loss occurring at that point is smaller. Second, the deflection of the flow at the transition from the windshield to the roof is weaker. The low pressure peak occurring there is therefore smaller and the positive pressure gradient in the remaining flow is less steep. Hence the momentum loss in the boundary layer is lower, allowing greater pressure recovery in the area of the rear end. . . . Therefore, even if a strongly inclined windshield does not contribute to a local drag reduction, it helps to improve the flow over the rear part of the car and thus to reduce the overall drag.