I have made headspace volume measurements similar to the ones
@doug293cz made, and the results agreed pretty well. The difference between keg and bottle headspace percentages (keeping in mind the fill level assumptions previously mentioned by
@doug293cz in this and other threads) is negligible.
The best evidence (besides basic physics) I can give for the idea of using the same amount of priming sugar for kegging as for bottling (keeping in mind the keg fill level assumptions previously mentioned by
@doug293cz) is that I have "keg primed" several times and actually measured the resulting pressure. At the end (when the pressure stopped changing), the measured pressure agreed closely with the pressures on the standard force carbonation charts for the temperature and the targeted volumes of CO2.