1. The "marketshare" that is referenced by Kantar is new sales marketshare, not overall marketshare. Its an important difference to point out.
2. Even with these gains in WP, iOS and Android are seeing similar absolute gains (e.g. in the 2-3% absolute marketshare growth). All WP is currently doing is taking a small piece of the symbian/BB pie
3. Given Microsoft's track record, the problem with WP is how long it takes for them to release a new product. Mango was what WP should have been initially, yet took a year to release. By the time WP8 is out, we will be looking at iOS6 and the Android version AFTER Jelly Bean.
Overall, I believe WP will not go away, but will stabilize at around 5-6% overall marketshare. The problem then is that given Microsofts development costs, they will continue to lose money. (Just look at Bing vs Google and even with 30% of US search shared, Bing loses billions of dollars a year)