Interesting project
Basically a low frequency phased-array radio telescope (like LOFAR down here on the ground). In order to process the data, the precise position of the cube-sats relative to each other must be known down to a small fraction of the wavelength used, but at low frequencies this should not be a huge challenge