Re: Yonah should be less than 1W and certainly no more than 2W, which means that during the "Load" test the Yonah was running at about 15W.
I doubt it. My guess is that Yonah ran at ~25W, which is a bit under the 31W TDP, and that includes the fact that the tested sku was bin n-1, at 2.0GHz. The idle test, IMO, probably lacked some power management modes, which is why it wasn't lower. For one thing, the tests were run on a desktop chipset, and also it was not production silicon, so some features may have still been disabled. In mobile conditions on production silicon, Yonah will dissipate around Dothan's idle power (~1-2W), according to Intel execs, but in these tests, it was clearly greater than this.