Realistically, not much. On a fundamental level, app store success and platform success aren't yet one and the same -- look at BlackBerry, for example -- but beyond that, there are several things to consider here. Android launched on the T-Mobile G1 alone, which means that for most of the period IDC mentions, the Android Market was really more of the "G1 Market" than anything else: one device, one carrier, one country. It wasn't until the first half of 2009 that additional markets and devices came online, and even then it was slow going -- typically one device on one carrier per market. Windows Phone 7's had a moderately more well-supported and well-rounded launch with devices from HTC, LG, Dell, and Samsung reaching multiple carriers in multiple countries across North America and Europe within a few days of each other. Sure, you could use the iPhone as the counterexample here; Apple saw explosive App Store growth with just two devices (the original and the 3G) on a handful of carriers around the world, but by the time third-party apps were enabled, the company had already assembled a big installed base of users hungry for more functionality.
And let's not forget Redmond's time-honored mantra: "developers, developers, developers." Microsoft is famously great at supporting its developer community, and it did a commendable job parlaying that developer base into its Windows Phone 7 launch by utilizing tools and languages that those developers already knew. Clearly, throwing together an app for a new platform is a much easier endeavor to justify when you don't need to crack any new books or learn any new development environments -- and indeed, an existing .NET or XNA developer can roll a WP7 app with very little ramp-up time. On the Android side, sure, the syntax is familiar to Java developers, but the libraries and constructs were all new; mix that in with an ultra-limited launch in the early months and you've got a recipe for a slow-growing Market.
Bottom line? Both these platforms are going to be successful, but for the Windows Marketplace to overtake the Android Market is going to take nothing short of a miracle -- and interpreting IDC's data any other way would be a reach.