When discussing how these algos work together, I always come back to ehow and other content farms that somehow survived the Panda purge. They fell briefly, but then recovered - why? They're still a content farm, still the very thing Panda was supposed to eliminate. Same with About.com. I assume the changes they made after falling were to barrage the algo and other computations with good signals that outweighed the "content farm" signals.
Does anyone know what they did to recover? What did they do that sent SUCH good signals, that Panda rolled over for them?