I'm not saying bulbs based on this research can never be mass produced, but it would require an extended development process to come up with an optimal commercial design, and manufacturing technology that does not yet exist. So don't be holding your breath on this.

In this work they are sandwiching an incandescent filament between plates of complex layered metal oxides, which reflect the heat back to the filament while letting the visible light through. Thus much less current is required to keep the filament hot enough to glow. The paper, being a letter rather than a full article, is rather sketchy about the actual structure and composition of the sandwiching plates. (See illustration below)

Though I have institutional access to the full paper, the abstract is available for free:
http://www.nature.com/nnano/journal/vaop/ncurrent/full/nnano.2015.309.html