This is a cool technology, making the assumption that it's real, but due to cost and scaling limitations, I doubt we'll see it outside of specialized use cases any time soon. For expensive use cases that have unique cases for it, it might prove interesting.