The difference between this idea and the EM drive is that there's an at least somewhat plausible theory behind it, that's not entirely incompatible with known s

The difference between this idea and the EM drive is that there's an at least somewhat plausible theory behind it, that's not entirely incompatible with known science. EM drive proponents could make the same statement and it would be less credible in that context. That's not to say that experimenting first and figuring out the theory later is a bad thing - that's often how science works - but generally, one follows strong evidence to reach a theory to explain it. EM drive proponents extrapolate wildly from extremely weak evidence. It's not good science.