I don't see the relevance of these issues to the role of dark matter. We see the light from a galaxy. It's from the past, of course, but we can determine its rotation at that time. We can also (given a good telescope) estimate the amount of mass contributed by the galaxy's visible stars. This is a snapshot of one moment, true, but at that moment the galaxy should behave in accordance with its mass, as we see that mass. But many galaxies behave as if there were much more mass involved. That's the situation. Thus the hypothesis of dark matter.
It wouldn't matter if, over time, the spiral arms contract, expand, or wave around. We see a galaxy's behavior at a specific time, and it appears not to have anywhere near enough visible mass to account for its behavior. That's the deal.
Yes it would, If the spiral arms were say in the process of expanding that would have an effect on the reading of the doppler shift. The problem making measurements with the spiral galaxies is they aren't symetrical and uniform so what is happening on one side isn't necessaraily happening on the other to balance it out. So when you compare the rotating away side with the rotating towards you side without knowing what else is going on the results will be off.