I assume so, but very general/handwavy with his wording, rather than on potential specific applications (of potential future mainstream use), which would help.
I think most people on this forum would be interested in how such can be related to gaming, but I know researchers that who be interested in cheaper/as or more accurate mapping (think "realtime" volumetric [mni-brain like] relative powermaps calc'd from beamforming/dsp on n-electrode arrays along the scalp at x sampling rates, and canceling out influences in the impedances due to rotational moments in the head [prob using gyros/3-axis accelerometers] and potential moisture build up along the scalp) for as much a 2-3 hours of scan time on a fmri/meg now, as well as being able to collect more data outside of the "lab" (and other brain states as well).
[0]: https://en.wikipedia.org/wiki/Brain–computer_interface