I'm gona try and not get too involved in this but will make a few points.
First off a hub figure is not a wheel figure anyway so cannot be directly compared to one. It is the power figure measured at the hub not including rotational mass of the wheel and tyre. This means it will be a higher figure than a wheel figure on a rolling road.
The TDI hub dyno is one of the only dynos that can be calibrated to make sure it's accurate, most others can't. Any dyno including the TDI hub dyno can give altered correction factors that can be controlled, it's upto the dyno operator to do his job properly and be fair.
They can all be manipulated by the air temp correction, the TDI dyno only gives hub figures and will not calculate (guess) flywheel but it does correct for air temp as does the dyno dynamics.
The SRR dyno dynamics is considered a very accurate dyno as the figures are very reserved and it's a great bench mark for us to compare. The problem with it and big power cars is they will not get traction.
The way the Dyno dynamics works is the car should be strapped in a way that it moves up out of the rollers slighty and on to the front ones. The problem is big power supras come on boost so hard and the rollers are so worn that there is just no grip.
The only way around this is to strap the car down so tight that it can't move, this helps traction but not only kills the tyres in about 2 seconds but saps alot the of power from the run. This is the reason we see such different readings from the 2 dynos with these cars.
Sam from TDI explains it the best way "a rolling road is like trying to measure a piston with a rubber caliper gauge, you will get a different measurement everytime depending on how hard you squeeze it together" it all depends on how it's strapped down, tyre compounds, temperatures and a lot of other factors.
So if we can strap a car down on a rolling road the same way everytime we should get some decent consistent readings but it will still never be as accurate as a hub dyno. The is the reason the race teams, the likes of shell and other big companies use hub dynos for testing. You can run it all day long and get very accurate readings.
I think one of the big problems here is we all got a little carried away with adding percentages on, a hub figure is exactly what it is and we should take it as that. Some of the hub dynos do read over but like I said its the responsibility of the operator and I could make one give out a high figure just by manipulating the intake temps. The only way to directly compare is to use the same dyno everytime.
We need to stop guessing the flywheel figures and just use wheel and hub figures, also we need a whip round so we can donate to SRR to buy some new rollers lol...
More importantly a dyno is a tuning tool and not an indication of how fast a car is. Get out there and enjoy them, how they drive is much more important than how much power they have.