It was actually published in January, and has just now been picked up by the Telegraph.
Here is the citation:
Ognjen Ilic, Peter Bermel, Gang Chen, John D. Joannopoulos, Ivan
Celanovic, Marin Soljačić. Tailoring high-temperature radiation and
the resurrection of the incandescent source. Nature Nanotechnology,
2016; DOI: 10.1038/nnano.2015.309
The Telegraph's explanation is terrible: "with a special crystal structure in the glass they can bounce back the energy which is usually lost in heat, while still allowing the light through."
The tungsten filament is sandwiched between two plates made up of layers of oxides, designed to selectively reflect infrared radiation and transmit visible light.
They used a numerical model to design and evaluate various candidate compositions for these plates. For a proof-of-concept, they chose one which uses layers of silicon oxide and tantalum oxide, with 90 layers in total per plate. This reflected about 90% of infrared radiation, producing a luminous efficiency of about 6.6%. This is comparable to commercial LEDs and compact fluorescents, though far from state-of-the-art.
However, the results closely matched their numerical model, and a more complex structure comprised of layers of silicon dioxide, aluminium oxide, tantalum oxide and titanium dioxide, with 300 layers in total, should produce a luminous efficiency of 40%. This is significantly better than the state-of-the-art in LEDs (about 15-30%). They did not, however, actually build that one.
I have no idea how expensive this would be to commercialise. It doesn't sound like the physics is particularly complex, but manufacturing costs could be prohibitive. I think it's safe to say we won't be seeing it outside the laboratory any time soon.