After reading the paper it seems that the biggest takeaway is that animal neural structures are sufficiently similar to human neural structures that they can process some of the same stimuli in much the same way.

The other interesting thing for me is the possibility that you could modulate these signals for RF transmission (I do like SDR's after all) and with a cochlear implant, and SDR, a computer to receive the signals from the radio and feed them to your ear, and a rodent wearing a backpack to send the data. You could have a rat running around the place and the "operator" could listen in on whatever the rat could hear. NOTE: There are at least two, if not four decimal orders of magnitude of work needed to get to something like that if it is possible. Not the least of which is increasing the bandwidth/fidelity of the hearing interface.