Picking up on the harmonization topic again, it appears that DIN has published an early draft of their report on translation between the two formats. The report outlines some preliminary findings in the feature differences between the format specs, and outlines some principles for what it thinks appropriate conversion scenarios are. It is great to see IQ being placed in this topic area, rather than the voodoo of wishing your neighbors cows will die J.

http://blogs.technet.com/photos/gray_knowlton/images/2992304/500x122.aspx

The report has examined a handful of functional areas to determine the level of feature parity between the specifications, and makes a comment about the "translatability" of that functionality. What is immediately evident is that there are feature differences between the formats. Nobody should be surprised by this. Looking at the table features identified in the report, we see examples of what the project is uncovering in terms of potential for fidelity.

http://blogs.technet.com/photos/gray_knowlton/images/2992306/482x480.aspx

Such a definitive guide will make it easier to translate between the two formats, and it will certainly inform applications seeking to improve interoperability. It will also underscore the idea that a "super-set" of all features in all authoring tools makes for an unwieldy standard. A "one (great big) size fits all" unified format is probably a bad idea; as we've seen with multiple file formats supported in countless applications, using the right format for the right task seems to be the preferred method (when to use TIFF vs. JPEG, for example).

Perhaps by design, there is no discussion about unification of these functionalities. For example, there is no discussion of a mixed and non-mixed content model unification, or a "merging" of things like shared formulas, graphic or table models between Open XML and ODF. 

The report seems to be feature-centric and focused on the ability to faithfully migrate between two formats. Perhaps the authors don't even bother to assume an interest level in "Unifying" those models, or perhaps this is scheduled for a future version of the report. (I would agree that Unification is not a good choice for harmonization.)

This is great progress; I am interested to see how the project develops.