Actually, they specifically took the weight out of the equation. They didn't start calculating the loss of fuel until after getting up to a certain speed. So I highly doubt it got better gas mileage from 0-60, since the added mass definitely took a toll. But keeping a vehicle at constant speed negates (for the most part) any weight, allowing them to measure the fuel efficiency based mainly on the shape of the car.
For those who didn't see the episode, here is what they did (in a little more detail).
They set up a long track with a start and end point. They accelerated a car way before the start point, and kept it at a constant speed from the start point to the end point, only calculating fuel efficiency from that point.
After that, they added a whole bunch of clay to the car. They shaped it to the car's mold, then did another test.
Then, they cut dimples out, making it look like a golf ball. They put the cut-out clay in a box in the car, so that it weighed the same. Then they tested the gas mileage.
The car with the dimples got better gas milage than both the smooth-clay car and the car before any clay.