I don't understand what happened in that stop-banging turn described above.
First, the left vs. right turn analysis seems to treat the disc as a fixed wing. Even so, given the small blade span on most gyros, the airspeed difference seen by the inner tip vs. the outer tip in a steeply banked level turn (owing to the difference in circumference traveled) won't be very big unless the overall airspeed and turn radius are very small. The overbanking tendency arising from such airspeed differences is rarely noticed in moderate span relatively fast airplanes but quite noteworthy in huge span slower sailplanes. For a VERY small span (rotor diameter) gyro, I think one would have to be already on the verge of a problem in level flight in order for the small speed difference from the roll-in to create a problem.
Further, reduced translational speed on the retreating side (inner, lower tip in a left turn for CCW rotor) from that circumference difference would not increase dissymmetry, would it? It's usually greater "backwards" flow on the retreating side that makes issues, and reduced wrong-way flow is helpful to compensation. Smaller translatioal speed subtracts less from the rotational airspeed seen by the retreating blade, and increases net airspeed.
One would also expect the g-load in a brisk turn to increase rotor rpm and provide margin against such problems.
I'm still troubled as to what really happened.