That's the average contribution from the electric motor during the whole test. However, that is NOT how these systems work - what they do is supply and absorb a lot more power for short bursts. I found the stats surprising - the batteries are very small, and as the first experiment indicated, they only use a small percentage of the batteries capacity. The reason for this is to extend the life of the battery - which is why when we talk about energy densities you have to be very careful - if you are only using 1/3 of the capacity of the battery, then its effective energy density is only 1/3 of the nominal figure. Flywheel systems run into a similar problem - by the time you are down to 50% rated speed it is very hard to extract the rest of the energy, although since that is only 25% of the nominal capacity it is not such a big deal.
Incidentally my spreadsheet predicts 55 mpg for the Prius - which just goes to show that with enough fudge factors you can prove anything. I need to sort out the driver model, then I'll post it up.
Cheers
Greg Locock