Most unipolars I've disassembled don't actually have alternating north/south... which might be better. E.G. North is usually on top, south at the bottom. Then, toothed, a bit like gears. OTOH, if viewed as a projection, I guess they do alternate. And the windings cover both halves...
I'm 50/50 on thinking this'd work. The other half says both motors would find a steady-state position they'd want to stay in, and trying to rotate one by hand would be resisted, just as if the motor was driven normally; the influence of one's slight flux-change would be minor in comparison to the winding currents. oh, and about 20% that I'm way off... so, 40/40/20, I guess.