I see two triangles, first; camera to grip to point on floor at tracks. The second triangle; camera to point on floor at tracks and then straight up to top of the crane (perpendicular to the floor's point). I believe (imho) that the math of the second triangle is to blame for the fall. As the distance between camera and the point on the arm of the crane expands out the force between the line from camera to track's point increases. I believe the solution would be more weight at the base of the crane to justify the extension of the crane's arm(with weight of camera operator)
While it's true extending out the grip might help, his position is more of a fluid up and down of the arm not stability of a crane.
I also agree that even an empty test run might (should) shown this instability where grips(human strength) might of caught the weight instability before catastrophic failure.