unfortunately many dyno plots leave a lot to be desired and tend to give you a rounded / smoothed result.... nothing like the actual curve. In this case the the original dyno only had a few plots (often between 10-20 points of reference over whole graph which is very coarse) which were then interpolated to make a pretty average curve.
We were able to give an accurate graph with far more refference plots (plots every 50 rpm on average) based actual time in that particular gear same as it would see on the road, this graph far closer represented the on road feel.
We did not touch the car tune wise at all, i dont think its fair on the other shop, im sure they will rectify it without fuss.
Luckily we are able to extract far more accurate results with the dynapck but we had to pay a premium for it $$$ wise when we first purchased the dyno, as for the power difference it is very common to see around 20rwkw difference from that particular dyno to mine, it is a genuinely low reading dyno compared to the various others in the area.
I know most dynos will read different but we are lucky enough to have at least five or 6 around that read within a very small margin which allows for quick easy comparisons. i even had the opportunity to directly compare a car from HPF on friday and the difference @310rwkw was 3 kw.... but our curve was much more detailed and showed up alot of detail not show on the other dyno plot... not bad tune detail but actual engine characteristics dulled out by roller inertia.
Dyno are tools and as long as you extract the relevant data your chasing in the most detail then the actual number is irrelevant and in this particular case a few simple runs showed exactly what needs rectifying.