The goal of these document formats should be to describe existing documents in an interoperable way, and to accommodate future enhancements or extensions. For better or worse, OOXML encapsulates, in an XML form, all the baggage of the Office file formats. Those have built up like a barnacle-encrusted hull over the past 25 years as Microsoft Office has evolved. ODF may be better in various ways, but it can only represent a subset of the features currently in Office.
The question I have is, why shouldn't OOXML be proven before being given the weight of an international standard? The Office binary formats are proven, yes, but OOXML itself still has a lot of fuzzy areas and question marks surrounding the details. Microsoft's open-source translator between Office binary formats and OOXML does seem to be making some progress, but it's not done yet. I'm concerned that in order to process OOXML effectively, non-Microsoft applications will be obligated to take on those 25 years of Office baggage and bloat.
In the midst of all this political wrangling about ODF and OOXML, let's keep the needs of users and businesses in mind. Microsoft shouldn't hijack this process, but it is practically essential that considerations about Microsoft Office document compatibility "hijack" the process, at least for OOXML. If ODF can't represent all the features of Office documents, then it won't be usable for two users that want to exchange Office documents. In that case, perhaps the two do need to stay separate, with OOXML looking backward and ODF looking forward.