I think that there is a lot of interesting things here, both about the specific case of Lagat's negative "B" (which implicitly means that there was a 'positive' "A", or the "B"test would not have occurred), and about the complexities of testing.
I wish that this material on Lagat was in a stand-alone thread with an appropriate title because I think it is worth saving.
spaniel, I appreciate your posting; I also think it useful to have at least some of the back-and-forth evident here. A lot of the comments raised by several individuals indicate a level of familiarity and claimed expertise such that in a typical thread the posters would 'carry' the censuses in misleading directions/results. The interplay here, while frustrating at times, highlights that discussion with links that help establish what is 'true'.