You assume fuel consumption corresponds to rpm - in reality, it depends on throttle position. You can be coasting downhill at 5,000 rpm, but if the throttle plate is closed, there's no air and thus no fuel. Cylinders are going up and down because the transmission is spinning the engine, but there's no fuel. When you put in the clutch, there's nothing to spin the engine, so you have to burn fuel to keep it from stalling.As many times as I've re-read that article, I don't understand how it can possibly consume more fuel at idle than it would when in gear and the rpm's are up higher. You would think lower rpm's less fuel, more rpm's more fuel is added. Plus, the ECU adjusts how much fuel is needed at certain rpm's. In the article they talk about pulse width and it going to 0, well then only time it really should be 0 is when the engine isn't running. Otherwise, it's 1 or higher. And in the article they state: Some small cars with roller cams may do slightly better. The article is almost 3 years old. The technology going into cars these days has me wonder what they have done to electrical, fuel and exhaust components to make cars get better fuel mileage and still deliver performance such as the ST with ecoboost.