Or rrocket could be wrong...
I know nothing about how dynos work, but it looks like rrocket's post on InsideLine got a response from the author. He does make a good point about the rear-end ratio: even if the transmission is 1:1, how would the dyno know if the car has a 2.78 or 3.73 rear end? Or maybe the rear-end ratio is something that the operator has to factor into the equation before the dyno spits out its number? Inquiring minds want to know...
From
http://blogs.insideline.com/straightline/2010/03/inside-line-dyno-tests-the-2011-ford-mustang-gt-50.html :
"The Dynojet itself doesn't care what gear the car's in. Remember, there's a final drive gear reduction after the transmission, which further alters the wheelspeed - engine speed relationship beyond what the transmission does, so there's nothing magical about 1:1.
In fact, dynoing in a gear that's 1:1 usually results in a hair less (yes, less) driveline loss as its a more (the most) efficient ratio in the gearbox. That's why a lot of dyno operators like the 1:1 ratio. Plus, when using an inertia dyno like a Dynojet, higher gear ratios have a lower rate of acceleration, so less power is soaked up in accelerating the rotating masses, which further props the numbers up.
Higher gears do tend to result in higher tire losses since the wheelspeed is higher... at this point all you're doing is trading off one loss mechanism for another. Generally, though, what I've seen come out in the wash is that on an inertia dyno, higher gears will result in higher numbers than lower gears.
Furthermore, higher gears also load the engine for a longer duration, requiring more cooldown, plus puts more heat stress on driven tires. As the dyno operator, I don't like either of those things."