If these are what I think they are, then one of my friends worked on them a couple years ago. The idea was basically stolen from up-and-coming LCD research in which a certain type of acid was used to lower the switching voltage. The acid essentially acts as a lubricant, decreasing the resistance to the crystal rotation. This resistance is what forces current LCDs to use such high voltages. By shrinking from a liquid crystal to a nanotube structure, the voltage can decrease much, much more using the same ideas.
I would guess that the single greatest problem is that scale-up of nanotube production is extremely complicated. A nanotube is essentially a polymer. When you produce a polymer, you pretty much always end up with a molecular weight distribution rather than a single molecular weight. However, for this sort of application, it would be imperative that all of the nanotubes be very, very similar in size. Thus, I'm guessing that development of the processes that create the nanotubes will be the step that makes or breaks the whole thing.
