[kach22i]

If I recall correctly, those roof wings were not true wings in that there was no flow under them, being sealed along the leading edge to the roof, more like roof spoilers they were.

Canted down a few degrees they induced quite a bit of lift.

The roof wing I made contributes to down force (see link in signature below).

The leading edge is beveled allowing an air slot or small opening at the leading edge.

The air in this slot I believe is accelerated by the narrowing, thereby lowering it's atmospheric pressure, and the higher pressure above the wing is drawn downward to fill the void left.

This then flows along the bed cover, where it meets the rear spoiler.

Not until this flow hits the spoiler is down-force realized.

At least this is my current theory.

The winglets on the Texas Tech roof spoilers delay vortexes from forming. Vortexes form at the intersection of side and roof because the air tends to follow a longer path along the roof and is at a different (lower) pressure than the side air. Once the flows meet and move into each other, they start to swirl, hence vortex formation and drag.