Good, so if we take for good the mentioned 20% efficiency of human muscles[1], the robotic arm wins largely, as 70W/0.2=350 W vs. the declared 200 W of the robo

Good, so if we take for good the mentioned 20% efficiency of human muscles[1], the robotic arm wins largely, as 70W/0.2=350 W vs. the declared 200 W of the robotic arm. (I still doubt that the robotic arm can be almost twice as efficient as a human arm, however, so probably the 20% is not accurate or there is some other factor involved) . [1] which I have no idea if correct even roughly