I was looking at this thread and with all the numbers being thrown around my level of suspicion (my distrust of behavioral studies statistics) started kicking in. ;-)
I saw Rrhainís figures based on a Washington march and remembered reading an article some time ago about the difficultly of establishing attendance numbers for public gatherings. Since I knew this was something that particularly the SFPD had been taken to task for (war protests, pride parades, etc), I thought perhaps there were some recent figures using more accurate methodologies. Unfortunately, I wasnít able to determine if theyíve been implemented as of yet. This seems to be a dead end, so I was racking my brain as to if there were any numbers which would be more trustworthy, when I remembered the census had started tracking non-married couples and specifically same-sex couples as of the last US Census.
I canít think of any reasonable ways to extrapolate the Census data to encompass the general population with any degree of certainty.
As to your figures, and I really might be way off here (having never taken any statistic courses):
Iím not sure where youíre getting the 7% figure, did you mean 0.7% and not 7%? Youíre referring to that earlier study you cited?
Iím not sure how one would go about making any comparisons between the given coupled same-sex populations and the unknown single same-sex populations. The societal presupposition is that same-sex male couples are less capable of forming and maintaining coupled-relationships. Even if that isnít the case, I canít think of a way to draw conclusions based on such dissimilar groups.