If you really want to test the part properly, don't do it at max boost and the limits of the turbo and the limits of your octane. Turn the boost down to 19 or whatever and run a tune that isn't close to the edge and doesn't knock at all. You might also want to flatten out all your timing tables so that it will run the same ignition timing every run. It will be hard to see the true gains of a specific part, if the turbo is maxed or your octane is maxed.
As far as Skunk2, welcome to the world of Honda throttle bodies. Leaks, not matching to the intake manifold, sticky throttles, idle problems, etc. I wouldn't be surprised if the 68mm TB didn't lose power in the midrange or not gain much up top. The 1G DSM TB has made 850awhp, Hondas make 500-600+ on their stock TBs. I am not saying that a bigger TB wouldn't help on a car like that, but it isn't a part that is known to be a big gainer on most cars.
I have been disappointed by the bored out TBs, ported intake manifolds, ported/coated exhaust parts on the Evos and Subarus. You are lucky to get 1/10th of the power gains that some of the people hawking these parts claim.
One other potential reason you might not see as big of gains is that you guys are trying to do an objective unbiased test. While some others will make sure to take the worst possible "before" run versus the best possible "after" run.
Scheides needs to put a Red on his car, rev it to 8500 and then some of these types of parts might actually start showing big gains over stock.
Sorry, if I am ranting.
Hopefully the Skunk2 TB is a good piece. I hate TBs that are too touchy or sticky.