A CdA of .21-.215 seems possible, and inline with Wiggin's past performances...on the track, and in a 10 with no/on turns. It seems completely unreasonable on a course like London. Indeed, the delta of your own CdA from "tunnel to road" is in that range of what, .8-1.5 (going on memory from an earlier post). And that's on a course like Moriarty? Really, almost any US tt course can have a reasonably low delta between straight road and event CdA. Most Euro courses are way more technical than that.
I have actually seen 1 power file from the Olympics, and it's right in line with the rider being about .02 higher than a "best" CdA.
Either way, we still come back to the point that he's performing significantly better to the competition compared to '09. This is pretty much unarguable if you look at his results. So, something's going on: everyone's going slower, he's increased his power or he's reduced the energy demands. The article which started all of this suggests that he's done the latter...by reducing his cadence. It's a pretty ridiculous claim, IMO. From a drag perspective I don't see a Tony Martin-like change in his position over the years. I actually don't see any change, not a surprise given his pedigree on the track. So, we' re left with the other two options.
You're right about one thing: according to the power data in the public domain, he hasn't increased his power. Given the long, documented history of TDF winners (and their respective camps) lying, you'll have to excuse me for being a bit distrustful and not accepting those numbers at face value.