Correct - that's the basis of my thesis. That "SF" isn't necessarily more NIMBYistic than sub-regions of other cities that have had "high" (say, 15,000 residents/sq mi or more) density for an extended period of time. If those other areas had experienced a similar historical accident of municipal lines (ie, only those parts of a Chicago that have exceeded 15,000+ residents per square mile since the 1950s were considered "Chicago"), the thesis is that SF and those sub regions wouldn't show
on average dramatically different levels of opposition to new development.
I guess another phrase this thesis is that SF's apparant NIMBYism is a misleading artifact of what you describe as a "historical accident of municipal lines."
This distortion, of course, can make SF look good as well as bad. You can pump SF up by talking about the ratio of :Resident compared to Houston, or Chicago, or LA, conveniently ignoring the fact that in Chicago, the residential suburbs are counted, whereas in SF, they aren't. My guess is that this ratio in LA or Chicago's densest 48 square miles would look quite similar to SF. Or, alternatively, you can make SF look bad by excluding the sort of developments that favor the burbs (such as percentage of households with children - when a family moves from SF to Redwood City, SF "loses" a family - in many areas, this would just be considered a move within the region).
Good or bad, my thesis is that these differences are largely an artifact of how the borders are drawn, rather than innate cultural differences between SF and other metropolitan areas. I don't doubt that these cultural difference exist (and they may in fact be driven by the fact that the electorate in SF isn't diluted by the votes in the surrounding areas), but I think they may be greatly overstated.