Here is an example that shows how the angular velocity of the racquet is very high as the racquet approaches the ball, but then slows down just before impact. Note how (in slo-mo time scale) that Gonzo's racquet rotates through ~90 degrees in the 2-second duration that it takes from the racquet to enter the field of view to when it impacts the ball (from the racquetface facing the screen, to an orientation perpendicular to the screen). And then during the next two seconds (mostly after impact), the racquet hardly rotates at all (it stays oriented perpendicular to the screen).
You can see the angular velocity by noticing how much the angle of the handle changes relative to your view between frames. I have looked through other players like Federer, Djokovic, and Soderling. And this pattern is the same for all of them.
Actually, the curves for omega2 in Fig2 of cross's paper (where C2 is set to zero) look more like the videos!
So I don't think there is a contradiction between my statements and Cross's simulations.