There's also some variants that do work, or at least could work in theory, which turn out to simply be highly inefficient photon drives. For instance, one recurring one that seems to get reinvented every few years is one where you put two electromagnets some distance from each other, and use the speed of light delay and some moderately clever polarity flipping to make it so the magnets always attract or repel each other in one direction, providing thrust. There's no reason why this won't work... except that also shoots huge amounts of defocused radio waves out the back, which is where the equal & opposite momentum is "coming from", and it's wildly less efficient in every conceivable way than simply shooting a conventional laser in the opposite of the direction you want to thrust. (And there's some other caveats too when you really get down to the engineering task of trying to flip the polarity of really powerful magnets at the necessary rate of speed. But it is conceivable that using real physics, you might be able to build something with this principle that could generate vanishing fractions of a Newton without necessarily blowing up....)

Based on the link you provide, and guessing what the QM equivalent of the differential sail (this thing, I think: [1]) would be, it seems to me this could fall into either class, something that simply won't work or something that will turn out to be another variant of inefficient electromagnetic radiation drive, once the full set of interactions is taken into account. (It's pretty easy to miss some of the more subtle ways of creating electromagnetic waves.)

[1]: http://www.daviddarling.info/encyclopedia/D/differential_sai...