given definitional problems, survey related limitations and significant underreporting, these figures are always going to be contentious, and it's probably a mistake to be doctrinaire about this figure or that.
as I mentioned upstream, I thought the RAINN figure was derived from NVCS figures alone; I was mistaken about that.
it's also clear that you minimized the degree of underreporting alluded to in the Sommers article.
I'll accept that these figures aren't cut and dried after looking at them a bit more carefully. Will you?