"any device with a thrust-to-power ratio greater than the photon rocket would be able to operate as a perpetual motion machine" https://arxiv.org/abs/1506.00494
Which raises the specualtive question of whether the Emdrive exploits the Casimir effect to obtain a tiny thrust (in which case a drive subdivided into a vast number of really small trapesoidal sections might produce much more thrust.
Exploiting the Casimir effect can't give free momentum, because the derivation of the force is rooted in quantum field theory, which like any other physical theory of the last 400 years contains conservation of momentum at its very core.
Any such effect that exists, would not be called the "Casimir effect" because it would have to obey completely different laws than the actual observed Casimir effect seems to. The Casimir effect is actually evidence against such an effect, insofar as it's perfectly explainable in terms of conservative forces.
I'm unsure about why you think that laws of conservation will not hold for Ether, AKA Quantum Field, AKA physical vacuum. If we will be able to get energy from QF, QF will lost that energy and will cool down, so Casimir effect will be smaller until it will vanish. However, there is no shortage of PV in space, and Sun is pretty good at heating it.
However, because of wave-particle duality, and because waves are propagating without moving, and because waves are propagating at speed of light, Casimir effect may be unaffected by rotation even when exposed to raw vacuum for speed less than 50% of speed of light.
The EMDrive breaks the known laws of physics as much as any potential perpetual motion machine would, so for it to be correct would imply as much is wrong as our knowledge of physics as the existence of a perpetual motion machine would.