Yes.
There isn't one properly isolated variable. There is more variables than just the wood in the test. Different bridge, different strings, human variation, ect. Properly Isolating variables is the basis for which an accurate scientific method study is done, and Rob's test doesn't do that, therefore it isn't properly following the scientific method. Forgive the elementary nature of the link, but here's an explanation since it appears the concept of different variables in the scientific method is going over your head.
http://www.sciencebuddies.org/science-fair-projects/project_variables.shtml
I don't need to have done a proper scientific study to know that Rob's isn't accurate. I personally have done a test with a replacement body, but I could not isolate the variables enough to conclude whether or not wood influences the amplified tone of an electric guitar. Until I see I strong scientific evidence to prove the popular theory of electric tone woods, I have no LOGICALLY reason to believe it. I used to believe it on anecdotes, however, like many people do, but that's simply not an accurate way to draw a conclusion.
No need to pinpoint. The tonal difference of picking position is a much easier isolated variable to test than two different wood bodies. If you pick the same guitar with different velocity and/or location, the sound alters drastically. That means that to test another variable, the picking variable must be isolated. I though this stuff was covered in high school, so hopefully you'll start recalling some of what we're talking about.
That's an anecdote. There's too many variables to consider (scale length, string type, string gauge, string age, ect, ect, ect) to conclude the differences/similarities is due to wood. Once again, you're entirely missing the whole concept of why a variable is isolated. You must isolate an independent variable to determine a dependent variable, NOT the other way around, which is what your talking about with your "fifth" point.
If his test isn't on a peer reviewable level, that its not accurate enough to draw any accurate conclusions from, and does not show the tone wood does or doesn't make a difference either way.
Well, you tried to. You failed to address the variables of the scientific method properly, and they weren't isolated properly in the test.
Once again, you have things backwards. You can't have a scientifically flawed test and say "one would need to prove" it wrong. A hypothesis isn't considered true or false until its proven so. The burden of scientific proof is on the one trying to prove a hypothesis, not people This test doesn't prove either because its a flawed test.