When I saw something similar as this over on "the other forum" and I stated a test fixture was necessary to remove the subjectivity from the test, I was reviled as a know nothing. Seems nothing much has changed. And that was many years ago.
I know how to set up an experiment. When a good experiment is set up and described, anyone anywhere can read my methodology, do what I did (exactly), and they should get the same result. If they do not, then they cry foul. Part of peer reviewed research.
You need a fixture to hold the machine, allow it to move up and down, and be adjustable (for weight), and something to hold the throttle wide open (no hand can touch this machine while cutting), and a way to let the platen drop into the wood in a manner that won't be violent (no...you still cannot touch it). You guys are smart, so think about it. A platen held in four uprights with Delrin bushings is a good start. Take into account the leverage the bar will have on the platen when resting in the wood so it moves up and down easily. The machine should be clamped tightly into the platen. Different bars weigh more and less, so adjust the weight of the platen so it is the same with each bar. The machine is tuned so it will enter the wood at the same rpm. Atmospheric conditions should be the same. The chain should be machine sharpened as no human can sharpen a chain exactly the same each time.
A good four foot section of tall tree, regardless of species, is pretty much consistent as long as it has no big knots. Now don't go off on fast growing pine and the same pine growing reeeaaallll slow in a bog, yes they will be different density, we know this. We choose a piece of wood that is grown in the same conditions as anywhere else, otherwise we get into testing for density of wood which is a hassle. I know wood in Australia is extremely hard, so using the same chain won't work as you need a chain made to cut efficiently in extremely hard wood. Choose wood that has the same density. This is going to be the tough variable to match region to region, nation to nation. The point here is: if your test set-up is the same as mine and if the density of the wood and diameter of log is pretty much the same, you'll get the same results as I do.
A camera with time stamp records the tach so you can see when the rpm drops and then, at the end of the cut, takes off. NOW we are ready.
The entire set-up is adjusted so the front of the machine is the same distance to the wood, make it one inch (your wood is held tighly by ratchet straps). The weight of the platen is the same, the saw is tuned the same, the atmospheric conditions are the same, the chains are ground the same, the wood is the same. Now...make a cookie. Saw not in the powerband? Add a lead ingot so it is drawn down right into the middle. Re-grind that chain. Now make three cookies. Look at the recordings of the tach, the drop in rpm and increase against the time stamp.
Repeat until testing complete.
AND I wasn't knocking what this guy did, only that the results cannot be considered seriously when a human is holding the saw. I admire the desire for knowledge and the time and effort he put into this.
I will not post another comment on this subject.