Ghurd,
His yaw mount is built like a dump truck. It looks too stiff to bend this much.
Using the yaw moment loads from Bergey's Excel (NREL test report TP-500-38550) to guess at the loads on Dan's, I'm going to assume about 10,000 inch-pounds moment (833 foot-pounds) have been applied either in a rapid yaw or furling event. With a 2" spindle (at the root) that is 6" long, then the tip of the spindle will bend only very slightly. The clearance between rotor and stator will change by less than 1/64".
Both yaw and furling pivot around the tower's axis. Since this is a big gyroscope, the force that acts on the hub is 90 degrees to the applied force. That means the action is to nod up or down, when the force is left-or-right. The stator would have scuffing at 12 o'clock or 6.
The windmill is free to yaw left or right, so vertical gusts do not apply much moment to the spindle before the windmill turns left or right, relieving the load.
Dan, I still don't know what could have happened to cause that scuffing. Didn't you say your windmill has over-sped and produced a lot of current? Is there any sign of heat damage in the stator? When the wire heats up the epoxy can flex, too. More like sag, actually. In the way you mounted the stator, is there a chance that the stator mounting bolts push the stator to bend? Though when cold the stator is stiff and the bend is negligible, when the stator is hot perhaps the pre-load can make the stator flex more.
Just ideas off the top of my head, because I don't believe it's a gyroscopic load problem.