I see the same difference with some of Wayne's tunes.
Torque is how much work is being done. Horsepower is how fast the work is being done.
With the graph I posted (I think from SteveRed's bike) the intake and exhausts aren't as modified as the guys running Unis and Jars. The modified intakes and exhaust allow the engine to develop power faster than the stock set up thus erasing some of the difference between torque and hp. Also part of the reason why sport bikes have triple digit hp and double digit torque.
The first thing I look at on a dyno graph is to verify that torque and hp cross at 5250 RPMs otherwise it is a bogus graph. Secondly look at which one offers the fattest torque curve. That is going to be the most efficient set up. Next you have to look at differences in intake and exhaust that would affect efficiency.
Also keep in mind that a PCIII tuned on a dynojet is having the computer tune it. It is probably the same program in each dynojet, thus they are going to be tuned in the same manner. With tuneboy, you have individual tuners that will tune differently.
Then you have difference in dynos, which is a whole 'nother dilemma in itself.