It seems to me that most of these drawbacks come from insisting on a monoculture, which -- if you have an efficient cellulose processing capacity -- I don't see why you care about. A real natural marsh or tallgrass ecosystem can't stand much mowing, but if you could tune a multispecies ecology for biofuel production, that would pose much less risk of invasive potential, and would be less trouble to look after. The tallgrass ecosystem was based on burning, anyway, so you should be able to maintain the soil while taking out hydrocarbons with minimal fertilizing.
Good insight, this. Cellulose is, for the most part, cellulose. At the point we commercialize cellulosic ethanol, it doesn't matter what the crop is. Just take whatever grows best locally and run with it.
I still think crop based biofuels are only a short term and/or small part of the solution. We can reduce our energy consumption many, many ways. We still need to eat. But if cellulose can be harvested from otherwise un-farmed land, maybe in lieu of prescribed burns, it'd be a win-win situation.