Preliminary numbers on the MCW6000; final results (w/ all the required reps) could shift up or down by 0.3C or so.
Bill's "2C" improvement seems pretty close to dead on. Block performs a bit better than I expected on my testbed in fact. Getting performance this good is a bit of a pita; the holddown has a lot of play and you have to get the block oriented "just right" as you are tightening it or temps are 1-1.5 degrees worse.
This is the new layout for graphs on the site btw (well almost). I'm going to play around in Origin today and see if I can't set up a simple template for dual Y axis graphs and put L/m on top as well