I don't buy the standard "open magnetic field lines" explanation for the fast solar wind through coronal holes, and for a number of reasons. First, plasma isn't really famous for strong magnetic dipoles. These are generally a consequence of stable electron orbitals, which plasma does not have. Ferromagnetism actually requires a crystal lattice, with orbitals that all get lined up and therefore contribute to the overall field, which obviously plasma does not have. But even if plasma had strong magnetic dipoles, "open field lines" wouldn't do anything. Magnetism can only accelerate dipoles where the magnetic lines of force are converging, and the acceleration is only in the direction of the convergence. For example, in the presence of a bar magnet, iron filings are attracted to either the N or the S pole, whichever is closer. So there is no way to accelerate a magnetic dipole away from one pole (going in a direction in which the lines of force are diverging) and all of the way to the other pole. The acceleration is only in the direction of convergence. Hence "open field lines" aren't going to accelerate anything, and diverging field lines will pull stuff back in (if the particles had strong magnetic dipoles, which they do not).
Aside from magnetized particles (which are not present), time-varying magnetic fields can induce electric currents. But this isn't relevant. The particles in the solar wind always flow away from the Sun. So if that's a current, it's a direct current. To generate a direct current, the magnetic field always has to vary in the same direction. If the field first goes one way and then the other, you get alternating current. So what kind of magnetic field would induce a steady direct current? Under the circumstances, there isn't one. So those are charged particles alright, but they aren't motivated by a magnetic flux.
That only leaves one possibility: the charged particles are motivated by an electric field between the Sun and the interplanetary medium. Alfven estimated this field at 1.6 GV, and that squares with other calculations I've made on the total number of watts generated by the flow of an electric current out of the Sun. To make a long story short, there is a steady stream of electrons emerging from the Sun, that light up the photosphere and the corona on their way out into the positively charged interplanetary medium.
So why would we see fast solar wind (i.e., 800 km/s) in the coronal holes, and slow solar wind (i.e., 400 km/s) in the helmet streamers? It seems that the density of ions in the corona varies. If the plasma is thick, the electrons streaming out of the Sun are slowed down as they light up the streamers. If the plasma is thinner, the electrons aren't slowed down at all, and they zip out into space at twice the speed.
Here's a graph that shows the inverse relationship between proton density (the red graph) and solar wind speed (the white graph):
http://qdl.scs-inc.us/2ndParty/Images/Charles/Sun/SolarWindSpeedAndDensity.png