First the link to this week’s complete list as HTML and as PDF.
Jasechko’s model stands and falls on his assumption, that plant transpiration does not fractionate. Not only is that theoretically highly improbable, but Kahmen et al. have shown fractionation to be the case and successfully used it as a climate indicator (PNAS 108 (2011), 1981-1986).
Kerig’s hypothesis is not new, only new to me, but like everything of his highly convincing.
Williams’ reconstruction may even be correct but he omits a simple and obvious confounder. As Boyd & Richerson have shown (American Antiquity 66 (2001), 387-411) humans can reproduce extremely rapidly if conditions are right. A pristine environment with no wildlife adaptation to humans and still extant megafauna was just that and the rise from the true founder population to his inferred higher one can easily have happened in an archaeologically invisibly short time span.