Well I read the article, all set to slag off the pseudo-science. But, to be fair, all he's done so far is define his terms. We don't even have a hypothesis to test yet.
I'd take issue with his argument that he must be unbiased because Banshee make bikes with all three wheel sizes though. He may be unbiased (although he'd be the first observer in history who was), but making bikes with all three wheel sizes doesn't mean that you don't have a preference for one. Maybe there is more profit in one. Maybe this is laying the groundwork so that they can drop one. I don't know, but Banshees current catalogue isn't really relevant.
I suspect that this article will just fall into the same trap as every other supposedly scientific study of wheel sizes. In an attempt to make it more scientific you try to just isolate the effect of the wheel and in the process make the whole thing irrelevant. In practice the wheel only matters within the context of the bike. For example, you may be able to show that wheel A rolls over a bump x% easier than wheel B, but without putting that into context it's useless. Wheel B may also roll over the bump x% easier if you reduce type pressure by a few psi, or change the head angle a fraction, or increase the travel my a few mm, or change any number of other parameters.
What matters is the whole package. A 120mm 29er with a 69 degree head angle and a 140mm 26" or 650B bike with a 67 degree head angle (for example) may both end up doing a very similar job as well as each other, just in a slightly different way.