Not sure why you think its not efficient in terms of motors. Think of it as a method of control, not a specific design and you'll realize it works with any number of motors.
I saw one like this at the national tournament, except the controller was a model of the robot with position sensors, not tied to the person's arm. Did OK, user needed more practice, a little better scoring strategy and a remote arm. His had trouble following the control inputs accurately.
From my observations at the national, users need to work on the robots stiffness for accuracy. Uncontrolled bouncing/overshooting makes it hard NOT to knock things over or out of bounds. I saw a lot of that. And on the robots speed. Three minutes is NOT a lot of time to complete the task. The cheap kits mentioned earlier on this thread besides being to short are also to SLOW. Students didn't have time to pick up more than a couple of objects and move them to the goal boxes.
Faruh's top score this year didn't seem to be the result of anything special in the robot (though it wasn't bad) but a result of LOTS of practice on the task with the robot and a laser like focus. We actually had trouble getting his attention when the five minute set up time was complete. Amazing concentration.