A few points to note:

Any inertia dyno will not provide readings that can be compared to the manufacturer's claims. If you take the time to actually read the SAE J1349 you will note that the power reading needs to be stabilized for 2 minutes for it to be accurate, and this includes all temperatures (EGT, Oil, Water, etc). You just can't do this on an inertia dyno and power levels will appear artificially high compared to OEM ratings due to this. Turbo/supercharged cars are even more susceptible to this.

Second, the graph at the top of the screen doesn't appear correct. The power curve should be less than the torque curve below 5252. Looks like the lines were mixed up before/after this point.

I have done a bit of research on different dynos and the techniques used to dyno and the ways you can intentionally or not, skew a dyno result.

Not that hard to fool a Dynojet or any Dyno for that matter.

I have however collected several dozen Dynojet runs from all over the country. I have also done interviews with the dyno operators/builders and techs of the dyno results that I see out of line from the norm. In most cases I can now easily pinpoint how the dynojet run was skewed very quickly.

All that said, at least with the engines I am most familiar with, I can pretty much tell you what the common mods will get for results. That is a proven over time an on a number of like engines.

The mods and the results are consistant enough to call a Dynojet machine a repeatable measurement for the majority. YMMV of course.

If you are getting really skewed results from a Dynojet I suspect your results are wrong elsewhere in my experince.

For example I have a couple of hot rod 911 engines dynoing 280hp or there abouts on a engine dyno. Put them on a Dynojet and they drop to 220 at the wheels....so I don't see that as giving overly optimistic results from a Dynojet. Just the opposite in fact.

A lot that can be done with dynos. Most importantly they are tuning tools and not a truly accurate measuring device because of the obvious variables on any given run. The lack in ability to make like comparisons is easily negated with a little effort and thoughtful comparison.

Bottom line? I'd trust a Dynojet and it's software over most anything else out there commonly available. Human error can be controled and limited with a little for thought and knowledge going in.

Ditto what Dane said. Before I jumped into the 3.6 process, I spent months analyzing different engine choices (remember, I had bought a 3.2 and was going to convert it to a twin plug 3.4?) --- as well as dynamometer runs from 3.2, 3.4, and then about 10 3.6 engines. Also searched www.Rennlist.com for many, many runs. Go on over and check it out.

I'd go out on a limb and say that Dynojet runs are pretty consistent from machine to machine -- and even quite fairly represent the 15% drivetrain loss 'rule'...

Ultimately, the same machine should be used for different engine comparisons - but with Ye Olde Dynojet 248C it seems to be ok to cross-compare data to other 248Cs.