As far as I can tell, everything is OK inside the servo. No missing teeth and everything engaging. The only weird thing, which means nothing since I know nothing about servo mechanics, is that when the main gear (the one that rotates the servo arm) is lifted out, the other gears actually all spin with the stick at neutral. This stops as soon as the main gear is replaced. Is this normal? I'm tested the arm "under load" so to speak and am not getting any slippage.

As I said everything looks and feels normal to me. And yet, I can't explain how the servo arm ended up in a completely different position after the crash landing. Could the crash have effected the servo's circuit board? Honestly, it wasn't that much of crash. The servo arm would have been "stressed" during the crash and that stress transferred to the gears but I don't see how that could have causes a glitch in the servo position. I certainly haven't been able to duplicate the glitch.