No, I`m still not impressed, reading this, they seem to be still limited to the melting point of tungsten?
Theres only so much heating you can do to a tungsten wire, (whether or not their reflecting more and more heat back to the filament over and over again), before it will physically melt, and i doubt thats anywhere near 100 lm/W???
I'm thinking they don't want to burn the filament hotter actually the reverse. Take a 40 watt lamp. It takes 40 watt to produce X lumens @ 1550 degrees C. If it is "recycling" the heat back to the filament to keep the filament at 1550 degrees C (same light output) will take less energy. The more heat retained in the lamp the less wasted though the glass.
Anyhow due to the lack of commercial intent, it will go nowhere. it sounds a great idea I think in that report it said the same thing was tried in the 80s but never took off.