KML was a de facto standard (at least in my book) before Google joined OGC. I've wondered a lot about why they bothered to join, and why they want to submit to a process that has a high degree of risk in terms of KML being mutated beyond recognition. Google's line has unwaveringly been that they want KML to be 'owned' by someone other than Google so that more people would be able to use it.

Since Google joined OGC, Microsoft has rejoined OGC and Microsoft has started to use KML in Virtual Earth. That's not to say that either event is directly attributable to KML being submitted to OGC.

Note, however, that OGC's blessing ought to lead eventually to ISO's blessing, and that could lead to far more extensive use of KML as a format for published data at the government level. Note also that NGA recently published the "National System for Geospatial-Intelligence (NSG) standards baseline" (http://www.nga.mil/NGASiteContent/StaticFiles/OCR/ncgis-eb.pdf) that mentions GML and a whole host of other OGC specs but does not mention KML.

Many government uses of KML exist. But they are probably mostly KML readers (i.e. Google Earth) or relatively small data sets.

If you were to compute the percentage of geospatial data held by governments worldwide vs. non-government data, I suspect the ratio is pretty largely on the government side. If all that data were to become available in formats Google prefers, that would be a big win for Google.