@old greg
BFSC under full throttle? I don't think people floor the gas when accelerating normally.
Also if you look closely at the scale, you'll see efficiency doesn't actually drop that much at higher rpm. Let's talk a little basic physics, I need a break from studying:
As rpm increases:
1. Friction loss from gears, piston ring/sidewall contact, is approximately constant.
2. Pumping loss increases at WOT, but this is superceded by the gains from 3.
3. "blowdown loss" (more charge in the cylinder, the more of it is wasted due to the expansion stroke not being long enough) decreases after a point because the intake system restriction becomes somewhat significant. With variable valve lift and timing this efficiency "increase" usually comes near the redline though, because the manufacturer wants to have as much max. power as possible. If you look at a typical BSFC chart, the best efficiency comes at about 75% max torque for an engine with a throttle plate. This is to say, filling the cylinder with too much air and fuel wastes more energy than directly giving up useful work to choke off some of that air supply. At higher rpm the maximum efficiency range moves up relative to max torque because of point 2.
4. Frictional loss in bearings around reciprocating parts (and other parasitic losses due to the vibration of the engine) increases as the square of engine speed. This is a somewhat big reason for decreased efficiency at higher rpm.
5. "relative ignition retardation loss" (I made up that term). I am referring to the fact that fuel burns relatively "slower" compared with piston movement as the engine speed increases. The rate that air/fuel mix burns in the cylinder doesn't increase as engine speed increases, but each stroke has less time to happen. Because of this we have to advance spark timing, but igniting the mix earlier and earlier causes the compression stroke to use more power to push against higher pressures for longer. I doubt this loss is that large because Formula 1 engines that run at ridiculous speeds still manage to run very efficiently by gasoline engine standards, although they have the advantage of higher compression, better balance and more advanced materials to reduce losses in other places. In a nutshell, thermodynamic cycle "improvements" don't see as much benefit as you'd think because energy is disappearing in larger quantities in so many other places.
5. Exhaust backpressure increases as the engine breathes air at a faster rate. Again this isn't a very big loss because compared to the other places energy is going, this is usually pretty small.
here is a chart that tells the full story:
http://www.pattakon.com/tempman/Saturn_BSFC.gif
As you can see, at the rev limit we are still within about 70% of maximum efficiency at full throttle. This is extremely good, considering that typically cruising along a highway runs your engine at less than 50% efficiency. Of course you have to take into account that manufacturers like to tune higher rpms for additional power output, which can be seen by the torque "jump" in the Honda F20C power chart, which represents a more aggressive and less efficient cam profile. I'm not saying revving high is good for fuel economy, it's clearly not, it's just not nearly as bad as what people make it out to be.