You have to remember that dynos (specifically chassis dynos) have great great tools, but they must be used properly. This is a very VERY common misunderstanding about them. You see it in virtually EVERY thread about dyno numbers on this board. This is separate from the issue of peak numbers meaning far less than the actual area under the curve (how the car is making power through the RPM band).
Properly used, dynos can be very precise tools. In other words, they have great "grouping". For instance, properly used, one specific BMW M5 may show 383-386 hp to the wheels, time and time again, as long as you're correcting for temperature, humidity, etc.
However, they're not very accurate tools. In other words, take the same BMW M5 to another properly used and maintained dyno, it it may show 420-425 hp to the wheels, time and time again.
Visualize this way:
High
accuracy, low precision (NOT NOT NOT how chassis dynos work):
Low
accuracy, high precision (this is how chassis dynos work):
This is why dynos are such great tools for measuring relative horsepower, but such absolutely poor tools at measuring absolute horsepower.
It's worth repeating:
Using a chassis dyno for absolute numbers (ie. I dyno'd 400hp, and I guestimate the drivetrain loss to be 15%, so my flywheel hp is 460) is ABSOLUTELY INCORRECT, and is a clear misunderstanding of how dynos should be used.
However, taking your car to one specific dyno and putting down 400hp, then making a change of some sort to your car (ie. exhaust) and putting down 410hp will allow you to reasonably accurately state that your exhaust change is netting you +10hp to the wheels, given your conditions (temperature, humidity, strapdown pressure, fan power & location, dyno operator procedures, etc) stay constant or are at least corrected for.
Here's more proof:
1 car, 6 different dynos (including 2 of the same manufacturer). The numbers are wildly different, eventhough we're talking about a car that's putting out just over 1/2 the HP our M5's are.
Turbo Magazine's Dyno Dash - Tech Review - Turbo Magazine
Here's another, dynojet numbers vs dynodynamics:
Dyno Comparison - Dynojet & DynoDynamics (Apples to Apples) - NASIOC
There are hundreds and hundreds of more examples.
Anyhow, while it is tempting to just compare peak chassis dyno numbers across different dynos, different cars, different situations, I hope we can all at least agree that this is using dynos improperly and the results of which do not actually mean anything.<!-- google_ad_section_end -->