A pretty poor report, constantly jumping back and forward from hypothesis to qualification of knowledge base. "Basically, our approach was to take imperfect but independent data sets and add them together," – but does that make the science twice as reliable, or twice as unreliable?