performance air dams often have a horozontal lip extending forward for a couple of inches. The purpose is to provide a surface for the high pressures area to act downward, improving the down force on the fornt of the car. This approach is probably only measurable at highers speeds, but in any event would not help mileage as it blocks more air and is designed to increase rolling resistance at high speeds.
One idea I've though about is to attempt to redirect more air upward over the better aerodynamic body shape, rather than under the car. A normal blunt nose would redirect air evenly. Perhaps a dam sloped up backward from the very bottom would redirect more air up...
But this is just a loose thought idea. I have not tested and usually things behave much differently than what we think, when it comes to aerodynamics... done properly you'd probably get some lift, which would be bad for traction, but maybe good for mileage (unless the lift had other negative altering effects like actually increasing air flow under the car... who knows.
The more I play with this stuff the more I see why the pros use wind tunnels and yarn tests. Even a small change can have unpredictable results (CFD programs just aint there yet for this kind of stuff, at least outside for us civilans