Seanw makes a great point about the TMIC and the airflow of the fan. I wouldn't trust any numbers that did not standardize the testing procedure. If you offer cooling air in a path perpendicular to the cooling apparatus you are, in effect, NOT offering cooling air at all. This engine is known to pull timing if charge temperature becomes high. You are, in essence, setting this engine up for disappointing results.
Now place a fan directly over top of the TMIC so that cool air can be directed into the path of the cooling apparatus and then we are talking a different situation. As mentioned, the TMIC, turbo-assisted vehicle has the most to lose by testing this way, the naturally-aspirated the least.
Given the limitations of conventional dyno testing, a more true test of engine performance might include a dyno test of a vehicle equipped with a FMIC MZR rather than the standard TMIC version so that cooling fans could be employed as they usually are without any modifications to the setup. Perhaps we could convince P5Freek to take his MS6 to the dyno to check it out, although the drivetrain loss would be greater since the MS6 is AWD rather than FWD as the MS3.
As a point of interest, I recall reading an article (I can't remember what online source it came from, but I will look it up and post the link when I find it) that new cars have sensors to tell when the car is actually moving forward with respect to the ground so that while the wheels may be turning on the dyno, the ECU recognizes that the car is NOT actually moving anywhere in space and thus does NOT offer up the maximum performance since the cooling fans CANNOT offer the same cooling effect as the natural airflow around the car when it is moving at speed. In the article cited, the authors test a BMW M5 on a dyno using several airflow configurations; no fan, a 24" shop-sized fan, a 48" industrial-sized fan, and another even larger super fan (the sizes are not exact, but rather relative from what I call of the article). They did several dyno runs with each size allowing the car to cool completely between runs so that engine starting temp did not play a role in the tests.
They found that no matter what they did, the M5 offered up only a finite number of HP on the dyno based on what the cooling configuration was. The stronger the cooling fan, and thus airflow presented to the engine via the front of the car, the more HP they were able to achieve. They were unable to get anywhere close to the claimed 505HP (after correction for drivetrain loss) with the small fan. As the size of the fan increased so too did the resultant HP at the dyno. This occured consistently as they changed the size of the fan. However, even with the largest fan they were not able to achieve the fabled 505HP.
In the end, their conclusion was that since the car was not actually moving at speed where the optimal cooling airflow could be achieved, the ECU was retarding the timing based on the amount of cooling air directed over the front of the car. They felt that with modern technology, tuners might have to develop ways of circumventing the ECU even in dyno testing so that the maximum potential of each engine could be achieved.
Just my humble $0.018 CAN...LOL
R