Many people believe that Mustang dynos read lower than DynoJet units -- comparing RWHP numbers measured on different dynos (even the same style of dyno) and under different conditions (ambient temps, humidity, etc) is like comparing apples to oranges. As as also been mentioned, the operator of the dyno can manipulate the numbers, typically to show greater gains from their product. As far as drivetrain loss to calculate BHP, 12-15% would probably get you close with a Lotus (it's a pretty efficient layout). All wheel drive vehicles will show a greater drivetrain loss, more like 20-25%. All of the numbers re: drivetrain loss you've heard are just guestimates.
A good explanation and I believe a good read of the Mustang vs DynoJet approach to measuring RWHP:
DynoJets are inertia dynos, and have been around for years, much longer than any type of load cell dyno. Inertia dyno's work on the principle of the acceleration of a known mass over time. Their rollers are the known mass. Weighing in at over 2500lbs or so. Your car gets strapped down to the machine, and the dyno collects it's data. It is able to calculate horsepower by measuring the acceleration in rpm of the rollers in regards to RPM. This is why gearing can affect the dyno results, more on that in a bit. Now that the dyno has recorded the horsepower curve, it can take the integral of that curve and get the torque curve. Since the dyno’s power calculations are based on the acceleration of mass over time in regards to RPM, gearing is very important. Since a vehicle with a lower gear ratio can accelerate the mass to a higher speed using less engine RPM, it will show a higher horsepower number than a car with a higher gear ratio. If a car is able to accelerate the dyno’s rollers from 200rpm (roller) to 300rpm (roller)in 1500rpm (engine), then the dyno is going to record more power than a car that did that in 2000rpm (engine).
Now we go to Mustang dyno’s and other loaded dyno’s. Our Mustang MD-1100SE dyno’s rollers weigh 2560lbs. That is the actual mass of the rollers, much like the DynoJet. That’s about where all the similarities end. When we get a car on our dyno, we enter two constants for the dyno’s algorithms. One being the vehicle weight, the other being what’s called “Horsepower At 50mph”. This is a number that represents how much horsepower it takes for the vehicle to push the air to maintain 50mph. This is used as the aerodynamic force. Mustang dyno’s are also equipped with a eddy currant load cell. Think of a magnetic brake from a freight train. This magnetic brake can apply enough resistance to stall a big rig. Off one side of the eddy currant load cell, there is a cantilever with a 5volt reference load sensor (strain gage). As the rollers are spinning this load sensor is measuring the actual torque being applied. So as the rollers spin, the load sensor is measuring the force being applied, sending that information to the dyno computer, taking into account the two constants entered earlier, computing the amount of resistance needed to be applied to the rollers to load the car so that the force of the rollers resistance is as close to the force the car sees on the street. The dyno is then able to calculate the total force being applied to the rollers in torque, and then taking the derivative of that torque curve to arrive at the horsepower curve. Since torque is an actual force of nature, like gravity and electricity, it can be directly measured. Horsepower is an idea that was thought up by man, and cannot be directly measured, only calculated.
I like to state it like this. . . I start by asking how much your car weighs, lets say 3500lbs. Now you take your car and you make a make a WOT rip in your tallest non overdrive gear, how much mass is your engine working against? 3500lbs right? Now you strap your car on a DynoJet and you make a WOT in the same gear, how much mass is your engine working against? 2500lbs right? Now you strap your car on a Mustang dyno, how much mass is your engineworking against? 2500lbs. Plus the resistance being applied by the eddy current generator. We’ve seen anywhere for 470lbs of resistance to over 700lbs of resistance as measured in PAU force in the data logs. So which one is more accurate? Well they their both accurate. If a DynoJet dyno says you made 460rwhp, then you made 460rwhp. If a Mustang dyno says you made 460rwhp, you also made 460rwhp. Now which one of those numbers best represents what your car is doing when its on the street. That’s a different question.
The most important thing to remember is that a dyno is a testing tool. If the numbers keep increasing, then you’re doing the right thing. We try to look over at NET gain, instead of Peak HP numbers. A 30rwhp increase is a 30rwhp increase regardless of what dyno it is on.
Now I can address how to calculate the difference between one type of dyno and another. Simply put, you can’t. Because Mustang dyno’s have so many more variables, it’s not a simple percentage difference. We’ve had cars that made 422rwhp on our Dyno, two days later make 458rwhp on a DynoJet the next day. We’ve also had cars that made 550rwhp on our dyno, make 650+rwhp on a DynoJet a few days later at another shops Dyno Day. For instance, my 2002 Z28 with a forged internal LS6 Heads/Cam/Intake, makes 460rwhp on our dyno. I thought that was a little low, since I’ve had cam only LS6 Z06 vettes make 450rwhp. So I overlaid the dyno graphs. Guess what, the PAU force for my car was almost 200lbs more than the C5Z06 that made 450rwhp with cam only. So I entered the weight and horsepower at 50 number for a C5Z06 and did another horsepower rip with my car. The only reason I did that was to compare Apples to Apples. This time my car made 490rwhp, no other changes. Now I don’t go around saying my car made 490rwhp, I say what it actually did with the correct information entered into the computer. It made 460rwhp. Now if I ever get a chance to take it on a DynoJet (which I plan to in the spring), I have no doubts it’ll be over 500rwhp. I know this based on airflow and fuel consumption on the data logs.
But since we’re asked this question constantly we're fairly conservative, and hence tell our customers that the difference is closer to 6-7%, but as you make more power, and the more your car weighs, the difference increases as well. You must remember, Dyno's regardless of the type are tuning tools, and are in no means meant to tell people how fast their car is. Now which one is more "real world" is a totally different question. I like to explain it like this..... If you drive your car in a situation in which you have no mass and you're in a vacuum, so basically if you do intergalactic racing in space, use a DynoJet. If your car sees gravity, and has an aerodynamic coefficient, and you race on a planet called Earth, then use a Mustang Dyno