Excellent point regarding the myology - articulation fidelity is often analyzed without incorporating constraints of muscular anatomy, including moment arm changes that may preclude the proposed motion.
Though, even without that, ramming the dorsal side of the femur against the lateral sides of the ilia, so the bones don't even articulate anymore, was pretty stupid.
And in this case it's not a mere moment-arm change. The pose shown in the talk would preclude the _existence_ of any kind of iliofemoralis muscle or the like, because the dorsal edge of the trochanteric crest (where such muscles attach) was rammed against the supracetabular crest or lack thereof.