I'm not analysing this from the perspective you think I am.
Our difference of opinion likely boils down to how much of a competitive advantage we think a native WP app would represent over an Android app running on WP.
My view is that if the app was good enough to be popular on Android, and a real demand exists for that app on WP, then I see no reason why most people won't be happy with the Android version. Making most people happy, on a platform that is unlikely to reach 15% market share in the U.S. anytime soon, is more than good enough for most companies. I could actually imagine many people preferring the Android version, as familiarity trumps usability every time (again, for most people).
If app X is the popular app on Android, I just don't see a lot of people choosing an alternative app on WP just because app X doesn't offer a native experience on WP. That is where I think you are going wrong. We here at WPC certainly would, but we are a minority.
It also seems to me that you are ignoring how most companies think. Most will ask themselves this:
Is it better to split my development budget of $200'000 between Android and WP, thereby improving both apps a little, or is it better to invest the entire budget into the Android app, thereby improving that app a lot, and then making that available on both platforms?
Given that choice, I'm quite certain almost all companies would prefer the later, particularly because you just can't afford to lose market leadership on Android. The Android battlefield is where you have to win your battles, and if you actually do have competition, it's far more important to invest the resources you have into improving your Android offerings. Diverting your resources from the most important battlefields is the best way to lose a war.
That's what I mean by binary. Android apps on WP is not a on/off switch. I guess MS can make a better job than BlackBerry, although I don't expect the Android apps running perfectly.
Many apps depend on Native Android APIs or Google services. There will be bugs and performance issues in the implementation of the runtime. Other problems in the integration with the operating system: Live tiles, notification center, sharing interfaces, copy paste, file selection., background task, etc.
So if they implement this thing we would have apps working perfectly, to app with perf problems, or bugs, or without integration with the OS, or with features that don't work, to apps that aren't compatible at all.
There is a great variety of apps too, not all apps are social networks that have to win in Android to be relevant. For example, the best buy app it doesn't matter if they win in Android they need the best experience to convert 'visits' in sales. Or a newspaper app needs engagement to make money, a me too app doesn't maximize the profit in many cases.
The design and esthetics is important in the consumer market. This factor make the difference in many cases. If the app break the predominant user experience it doesn't feel good, natural. Or it could have usability problems if the behavior is different compared to the apps where the user spend most of the time. A developer has just 30 seconds or a minute to capture the user before she just press the home button and forget the app forever.
The size of the user base also make a big difference because the revenue depends on the number of users and the quality of the users.
So, it's not a on/off situation. The decision to go native/android/html5 would depends on multiple variables that are changing all the time in a continuous range. This create an app store in which the three kind of app coexist.