[elhigh]

In order to avoid further snafus like Lugnut's numbers getting thrown out because nobody trusted the pump (!), maybe the next time this thing is run, it should be done using independent meters.

Entrant arrives with full tank and steps out of his car. Onsite volunteers step up and seal everything - hood, trunk, fuel cap etc. with tamper-evident seals. Car is weighed with no-one aboard. Taking a couple of pictures of the interior prior to releasing the vehicle to the entrant would be a good idea.

Entrant runs the event. Seals are checked at the waypoint. If any are missing, the car is removed from competition. The waypoint gives the entrant the opportunity to procure needed items without pulling the seals.

Upon returning to the starting line, the vehicle is parked and the seals checked. Interior photographs are checked and if necessary foreign objects introduced to the car's interior along the way are removed to prevent throwing off the weight. Now the car is weighed again. The missing weight is returned to the car as fuel, and the necessary volume is noted as the car's fuel usage for the route.

I think that would do it. It's a lot more hands-on, requiring that there be bulk gasoline on hand but I think that's a lot more reliable than the pump's first click. And there are so many eyes working over all the cars I think it'd be hard for a cheater to get much done.

The photographs are necessary. In order to throw off his results higher, a cheater would have to make his car gain weight between the start and stop. It'd be awfully funny for a cheater to rig a trapdoor or something, shed a few pounds of water ballast on the road, and then get back and find out his calculated mpg is way below the EPA - and they've overfilled his gas tank!