The analysis you posted is very good because it tries to explain the larger picture.

"Microsoft doesn't have an agile management structure. It's heavily layered."

Indeed, that seems to be their problem!

The other side is that sales dictates and the company lives very well with that dictatorship milking its cash cows. Commercially it is stronger than ever. But for a technology company it is poison to work with all the PR spin and layers of worthless documents.

And in public policy they hire ruthless hitmen and lose with politicians. When you hire the tabacco lobby you will get the same support the tabacco industry got for its agenda. So you have to use your Microsoft mail client? Go to the small computer room at the end of the corridor. It is politicians now which talk about their plans to get rid off Microsoft.

Not to mentions that they don't "inspire" developers anymore despite that many of their products are really good horses. I talked to so many former users of Microsoft and they always feel like ripped off by that company and prefer to use semi-ready open source products, and they are really passionate about that. I can't be the alternative products.

I am very sceptical about premature funeral preperations. It is an important company that it going to stay in the market for long.

http://channel9.msdn.com/ demonstrates that they are able to learn.

"And those web apps play into Microsoft's strategy for capturing the Office 2.0 web app market and responding to that threat, including Google."

It is a kind of bubble, a conservative strategy is just fine. Recently Adobe bought their little toy. Microsoft's mistake here is that they took it far to serious and heavily invest in their live model. Google's real challenge is its agile development structure with which they were able to make inspiring innovative software and open spaces for really strong business models. Google Doc is just one of these tools but how many developers and strategists did they need to come up with that? How many developers build a service as wikidot? For the new round of Web business you need a kind of O'Reilly Summer camp with gifted imaginative developers who just want to build something useful and are paid to work "on something". And you can be sure that you can harvest at the end of the round. But you should not manipulate the development process by business expectations because that won't work.

The difference between science/research and agile is that science is not production focussed and adheres to certain reporting standards and assigned roles. In knowledge markets self-selection works better. Management always tried to punish the lamers and increase the medium productivity, thus framing the more productive persons as well. It is a risk-aversion strategy. But as in financial market more volatility is a very good thing. All organisations become orthodox and get their arteries sclerosis, the lawyers and technocrats take over, skills are replaced by qualification, so they need to implement agile methods and reinvent themselves. Anti-aging tools, get back to play with toys. Old FFII was very successful with its Artus court lobbying style. We are getting back to that productivity.

"Two-way high fidelity interop for competitors is anathema to Microsoft's present business model. The company has been built around interoperability of its own stack whilst denying interoperability to competitors."

Indeed, when we compare Microsoft to Apple's platforms it is an interoperability paradise. But as of open source they got trapped into expectations of the public at large and sticked to ideology. Microsoft was able to deliver a Linux IE client, no big deal. Why didn't they do? It does not even work with Wine.
http://appdb.winehq.org/objectManager.php?sClass=version&iId=4195
etc.

All Quiet On The Western Interoperability - Anti Trust Front