The device, without the counterweight and projectile, must not contribute energy to the launch. (rule 3d). I think all "walking arm" designs I've seen would fail this automatically. It would also be challenging to launch such a device from 75cm outside the launch area.
Perhaps you could construct a such a device where the only loss of gravitational potential energy between the beginning and end of the launch cycle was in the counterweight, but I think such a device would lose many of the advantages of the walking arm design. But if you think it's viable, it might be fun to build a small tabletop model and see how well it worked.
(Also remember that the goal of this event is accuracy and repeatability rather than just range.)
