Pretty good article, but a couple additional thoughts.

First, helping people leverage existing skills like .NET *IS* one of the key value props for Windows Azure, but there are a number of other important differentiators, as well. For example:

1. You can leverage your existing on-premise identity management infrastructure to secure app running on Windows Azure (and do it with industry standard approach - SAML tokens)

2. Today you can use a variety of languages via .NET - but also in the future, MS said at PDC, non-.NET would also run on Azure.

3. Portability. You can run build and run your app in Windows Azure, and if you're needs change, you can move the code to your own Data Center and run it there. (From CTP, it probably won't be zero work, but also not huge work).

4. Management - Windows Azure will let you manage your apps running there from the same Systems Center environment that manages your on-premise stuff.

5. Windows Azure is not designed to be .NET in the cloud but Windows in the cloud. That is, SharePoint Services will be available for building collaborative apps, CRM Services will be available for building custom CRM-like apps (in addition to the packaged CRM Online), Live Services is available for building social apps that help you connect with the 460M active Windows Live ID users -- not to mention create contextually aware experiences across browser, PC, and phones (and even non-traditional devices like cars, picture frames, TVs, etc). According to news this week, this is how BlockBuster plans to let you download movie at an airport kiosk, watch half on the plane, and watch the rest of it on your TV at at home starting from where you left off.

The second point is that what Microsoft is offering does not just differentiate on a feature-list basis, but it's fundamentally different in it's intent. That is, microsoft's goal with Windows Azure is NOT to help people build everything in the cloud but to evolve how you think about applications.

In a nutshell, Windows Azure (and other complementary moves like MS Online offerings) enable you to re-think what bits of functionality should run where -- on clients/phones/devices, on your servers, with hosters, and in the cloud -- so that you can better optimize the balance of user experience, regulatory requirements, and economics. And even change what goes where over time as your needs change.

In practice, some apps won't be in the cloud for a long time, some will be soon, some will move back and forth -- and some will be hybrids with pieces running all over the place... and yet still secure and interoperable with all your legacy systems.