Different dyno's do read different figures often
Same dyno's can read different figures on different days
road data logging is not live mapping, its logging.
RR is safer, strapped down doing 100mph safely. (no traffic)
depending on applied loads and dyno control, lower powered cars (like more std level ones) can and do often read standard power and the proclamation of it must be accurate is made. then when higher powered, higher boost cars are run, and make less than expected results. RR to blame? Car to blame? map to blame? at this point people want to "blame" something.. throw a dice and place your bets folks
the reason for many dyno's reading higher powered cars lower than relatively standard ones, especially on the load based ecu's like Me7's, is the runs do not simulate acceleration run rates to that of on the road. Often is an rpm per second ramp rate fixed. rough
guide is know what your on the road acceleration time is in the chosen same test gear, and compare that to the applied brake on the dyno (if the dyno has eddy brake and is'nt just fixed mass inertia dyno) - it will give you a 'feel' for relative comparison. Also logging during the runs is well advised, so you can see and compare dyno to road..
relative comparisons on the day on similarly spec'd cars is about as best as you will get.
got to 5 different dyno's & you will probably get 5 different results, in both terms of maximums recorded as well as how they are delivered in terms of the dyno plots ( shape of power and torque delivery etc etc)
RR is a good tool.. I cant wait to get my own.