I think the concept with the "discharge the cloud" approach is that you put up a whole lot of lightning rods over the area to be protected. That is supposed to then provide a path for current flow between the cloud and ground that in turn lowers the potential difference. The obvious question is how much current can actually flow compared to the enormous capacity of the cloud? During a lightning strike you have an ionized path which provides good conductivity. And there is obviously a huge amount of energy to discharge. With just lightning rods and no actual strike to provide ionization my bet is that any discharge is going to be negligible and hence the rods would just serve as a lot of conventional rods.