Maybe I'm reading the article wrong, but it seems they are only describing a way to record EMG signals --- and not a way to `control' anything. I've previously worked in neural/EMG recording of flying insects and a major issue to overcome is the effect of the recording apparatus on normal behavior. For instance, a tiny coin-cell battery can have a significant impact on the flying ability of a small insect (we targeted dragonflies). This work described in the article seems to use a very large structure attached to the insect which the insect cannot lift (they describe being held aloft with electromagnetics). I'm definitely curious to see where they go with this though. I'd love to get back into 'insect cyborg' work.
EDIT: And after looking a bit more, it seems that the actual paper describes the surgical process for implanting the electrodes such that the rebuilding of the moths tissue actually makes them be a part of the moth. Cool!