Didn't they use this for siRNA transfection in their publication? I haven't spend a lot of time looking at their exact setup, but the concept behind their model doesn't seem impossible to adopt into a mass-producible product that would be expense, but not prohibitive. While a lot more physiology needs to be studied before we can understand what kinds of drawbacks this might make, current methods of changing lipid composition or poking electrically-induced holes in the membrane (without a needle to fill it) have significant changes in the membrane structure. If you're studying membrane proteins/interactions this is a significant drawback. At least in this technique, you have some knowledge of exactly how big and numerous these needle-induced holes are.

This really wouldn't have an potential outside of the lab in terms of pathogenic entry. If you used the technique to inject material into cells that were designed for later human implantation, they would have been transferred to non-spiked surfaces for at least sometime after molecule injection and before implantation. Thus the pathogen's entry point would have been long severed.

And how! But actually, a bunch of tiny pricks or a big shock of electricity? Hmm.

As someone who has spent plenty of hours in lab begging my cells to take up whatever GFP protein is the flavor of the week, something like this really could be interesting. As I see it, this would be a whole new class of transfection protocols in addition to chemical and electrical methods. Cost and the idea of actually poking holes makes it more similar to the latter, but it does have some unique differences. The most obvious is that you'd have a broader class of molecules that one can inject since there is practically no membrane interaction. Also, while the plates may be costly, there is no need for an expensive electroporation machine.