Actually, the opposite has happened. Much of the Android sales are in the high end, and every Android handset maker has actually increased margins. Well maybe not for LG, whose main Android seller is between low to mid end.
The risk of commoditization is higher on WP7 for the reason because there are strict minimum requirements and even the CPU is stated: Qualcomm 8250 Snapdragon, as you can see with all the WP7 handsets. I don't even see the leverage to use different CPUs. The result is that everyone ships phones meeting only this minimum spec or just slightly above it (8mp camera instead of 5, or 16gb instead of 8gb). Since you have 10 phone models competing on the same narrow price and market range, the result is product dilution leading to rapid price and margin drops. Ergo, you go BOGO so soon after release.
In any case, commoditization is really a product of supply vs. demand, along with careful matching of the Moore's Cycle (doubling power every 18 months). If you don't release a device that matches the curve in Moore's cycle, someone else will, making your product obsolete (your Nokia and Blackberry on ARM11s while LGs are on dual core A9s.) The high end buyers are the bleeding edge geek buyers, you want high end, you sell to the geeks to whom the specs matter. The farther you are from Moore's curve, the quicker you have to downprice because as you can see with the computer industry, the demand and margins are highest on the latest, fastest, bleeding edge products, and this moving curve is the Ground Zero for margins.
We are just days from ending 2010. 2011 has been proclaimed by some as the age of the dual core mobiles, maybe the year of the nVidia Tegra 2, if 2010 was the year of the Qualcomm Snapdragon.
The problem I see with Nokia---and Microsoft---is the way they set targets. Fighter pilots would call it "Lag targeting" --- targeting already what is behind. Even "prime targeting" won't work here. You target your target dead center, and by the time the bullets fly to reach it, it would have fallen way behind the tail. What you need is "lead targeting". You set your sights at the target several steps ahead, so when the bullets fly, it falls right in the center. This is where both Nokia and Microsoft has been seriously failing.
If Nokia needs to go "lead targeting", forget about WP7, forget even about Android. That's just lag or prime targeting now. If that was back in 2009, I would have advised Nokia to go Android. But not at the end of 2010. WP7 won't help for that matter given how behind it is on the feature level with iOS and Android and Microsoft has a slow rate of system updates.
The proper "lead targeting" strategy for Nokia is to go dual or multiple core tablets and smartphones, use Meego for such. Nokia needs to exploit its relationship with Intel, not Microsoft. Intel is looking for a lead to break into the smartphone and tablet market, and trying to gear the next generation of Atom processors to deliver. Concentrate on apps, no doubt, but one needs to concentrate best on HTML5 webapps, as that will be the future of mobile. Hence, the Webkit browser will be crucial here.
Just saying this. If you're a strategist, following the conventional and known strategies will only get you beat. I have to say that Blackberry has something about that new Onyx OS of its for that tablet.