While enabling both cores and running Prime 95 for the fourth test run, a 2.4GHz Toledo core managed to hit 86W. I imagine a 2.0GHz core would dissipate less, especially given AMD's rather steep power curve in the 2.0-2.6GHz frequency range. It's remarkable, for example, that their 2.2GHz Venice core only dissipates 31.6W in this test, while the same stepping San Diego at 2.8GHz dissipates almost twice as much at 60.4W. It suggests that a 2.8GHz Toledo core at this stepping would dissipate exceedingly high power.
At any rate, the Anandtech numbers seem in line with what I'd expect from Yonah and a 2.0GHz Toledo core.