Grr. I can't access that paper but thank you for digging that up. If Animats' calculations above are accurate though then uBeam is at a very small fraction of that 53%.
Is there anything in that paper that could explain the difference?
Is there anything in that paper that could explain the difference?
I don't know any more because that was how I found that and skimmed the paper about three years ago, for a prior discussion of uBeam here on HN.