I believe it is roughly 10%-15% higher than whats at the wheel on a completley STOCK setup .
If I had to guess it is based on the torque target number that the ECU is constantly calculating for wastegate control purposes . That number is an extrapolation based on all sorts of data such as boost , injector flow , ect .
When you consider that the ECUs calculations are probably meant to reflect flywheel HP , then the 10% or so error on the high side makes sense .
I think that most interceptors have a correction factor setting in the gauge that will allow you to syncronize your readings to reality if you wish . This can be accomplished by comparing actual dyno results to gauge readings . if the dyno says you are making 300 , but the gauge says you are running 330 , then set the correction factor to 0.91 . at that point your guage may not be absolutely perfect , but it will be very close to actual HP in most of the range .
Of course , changing things like volumetric efficiency (porting), bigger injectors , different turbo , map clamps/piggybacks will all have an effect on the calculations accuracy .
Even though my gauge is not spot on . it is still very usefull data in a relativistic way , that is , if I make a change in my tune I can see the difference between the before & after pulls . If I see a 10 HP increase , I know that may not be the exact increase , but I know Im on the right track . If I make a change that nets me nothing , or reduces HP reading by 2-3 HP , then I know not to go that direction . Get what I mean ? Its a tool , not a certified mini-dyno .
__________________
_____  _____ 
|