Loads of us here go on about the rolling road figures for our cars, then we get the usual 'Oh they're not accurate, there's no way you're getting that much from that engine!'.
What is classed as an accurate rolling road? If people have had their cars RR'd on a few different roads, and they all come back with similar figures, then is this not good enough as a test?
If a group of cars are tested and, where they have been done elsewhere before, have similar figures to previously, then again is this not acceptable to then trust the figures given to cars which have never been tested before?