Dembski's method is wrong because its trying to calculate the flagellum forming just by random chance alone and doesn't account for selective pressure so it isn't ruling out evolution.
And the working specification is wrong because it assumes that all function was lost.
SO is wrong about not being able to determine info about a designer from its design. He thinks that we'd have to be able to determine the name of the factory worker who stitched the soccer ball to know anything at all about the designer when we would be able to some things about the designer, and their mechanism too, by looking at the ball.