As far as I can gleam from this report, they want to revitalize the old "bent pipe" model of satellite communication, where the sat only acts as a repeater . I am still very wary of it being feasible. While the track of BS->MS is feasible, albeit with a high transmission loss, i cannot fathom how they want to receive the <1W signal of the MS. Conventional way to pick up a signal below the noise level would be decorrelation, and while 3G/4G work on that principle, there would have to be additional information exchange between BS and Satellite. Additionally, for that to work the SATs have to be in LEO, and as they describe them as "massive"(most probably the high-gain antennas), there then will be massive atmospheric drag, requiring either frequent reboosting or replacing.