Actually existing electrostat efficiency is actually very high: very, very high in fact, because apart from very, very modest losses in any necessary step-up transformer, the unit itself is essentially pure capacitance - and the dissipation in a pure capacitance tends to zero. Added to which the impedance match - effectiveness of coupling - between existing diaphargm materials and the air load is excellent - the major reason for the outstanding performance of good electrostats. [ Also why box speaker are so ineffcient, and exactly how horns gain in efficacy.]
It takes a wide voltage swing to drive adequately (hence apparent voltage insentivity /need for 50+w amp) but efficiency for electrostats is certainly not a problem. Th truth is, they'd benefit from a different approach to amplification - massive voltage swing at low current - and the practicality of that is not universal/plain dangerous in the home (sell a small amp that can easily swing over +/-3KV? good luck with that; what connectors would you use for speaker cable, in fact, what cable for a start...)
Added to which - macro-scale electrostats need very, very high sheet resistivity to approximate the constant-charge criterion which was the real Walker/Baxandall breakthrough. Making electrostats out of low-resistance stuff, however thin, is a recipe for much higher distortion and poorer perfromance. Strange but true: useful membranes start above 1gigaohm/square of resistance - nearly a perfect insulator!
The linked article is clearly using tiy pieces of minimum feasible mass to chase a Medium-wave frequency response: alone, that's of no use to audio replay. I'll stick my neck out: graphene does not intrinsically, even potentially, offer electrostat loudspeakers much if anything at all. Doubtless we'll duly see it marketed as the saviour of the audio world though...
