A fatal flaw in your analysis
Your analysis contains a fatal flaw, Martin. That is your belief that adequate Microsoft XML <> OpenDocument translators will be available. In fact, all of the translators suck mightily and there is no prospect at all of them being perfected.
The major problems are: (i) that Microsoft's XML formats seem deliberately designed to thwart their parsing with XPath, which is essential to XML transformations; (ii) that Microsoft's "XML" file formats include binary blobs, bitmasks, and multiple Windows and Microsoft dependendencies, all of which defy XML transformations; and (iii) OpenDocument assumes a richer page layout engine than Microsoft Word provides, so while DOCX can be completely mapped to ODT it is impossible to fully map in the other direction without declaring an MS Office interoperability subset of OpenDocument and ODF applications implementing a compatibility mode with reduced features. (That is more than somewhat ironic, given Microsoft's spin that it couldn't implement all of its features in OpenDocument. In fact, the exact opposite is true.)
In fact, Steve Ballmer is on record as saying that the developers of the Novell-Microsoft-Clever Age plug-ins will not even attempt to achieve full fidelity file translations between the two formats. http://www.eweek.com/article2/0,1895,2050848,00.asp?kc=EWEWEMNL103006EP17A
Those translators achieve at best far less conversion fidelity than existing file conversion filters between OpenDocument and Microsoft binary file formats such as the OpenOffice.org conversion filters, which achieve only about 80 per cent fidelity. The file format cognescenti know this. See e.g., the paper by Gary Edwards and Sam Hiser included in this edition of the European Journal for the Informatics Professional. http://www.upgrade-cepis.org/issues/2006/6/up7-6Hiser.pdf (PDF). (Note that I contributed to that paper.)
And as also detailed in that paper, what works well enough for some of us does not necessarily work well enough for all. Anything less than full fidelity data conversions is absolutely unacceptable in the context of wholly automated business processes and is in fact illegal in various contexts, including government records.
So your thesis doesn't fly. In fact, I'd go so far as to bet that you have been suckered by the Microsoft spin doctors. Another indication is your depiction of the file format wars as being waged primarily between IBM and Microsoft, a recent theme of Microsoft's public relations machine. While it is seductive to believe that the controversy is just another chapter in the war between major competitors, the pro-ODF camp is far broader than IBM.
For example, nearly 20 governments recently opposed fast track processing of Microsoft's draft standard at ISO. Do you believe they were all carrying water for IBM? Government bodies in more than 50 nations have chosen to adopt ODF. http://opendocumentfellowship.org/government/precedent And dozens of developers now support the OpenDocument standard in their applications. http://opendocumentfellowship.org/applications
While IBM has had a noteworthy role in proliferating the OpenDocument formats, there is a movement without a recognizable leader in the industry. When it comes to vendor influence on things relevant to ODF, Sun Microsystem's far outshines IBM. But in fact, a core group of open standards and free and open source developers and advocates -- inside and outside government -- have played a far larger role. This is a customer-driven phenomenon, not a vendor-driven effort as you portray.
So I will respectfully suggest that you reexamine your position on these issues. Reasonable minds can differ, but not on the grounds you advocate.