This weekend I realised that by necessity, it will need a reasonably large aperture to harvest enough energy, which explains to a large degree the size of the brick attached to the phone.
Even if all the other planets aligned and the other practicalities and regulatory issues were dealt with, increasing the footprint of the phone sufficiently for the aperture, and having to have the device oriented towards the energy source enough to make it "charge" (i.e., face down, and nothing covering the aperture), alone makes this application a non-starter.
Anyone could see that from day 1.
People put their phones down flat on the bench most of the time. You know, right were you put a $5 Qi charging pad that is vastly more efficient and cheaper.
There were two important additional facets that affect user acceptance, and I hadn't fully realized until now, and I'm not at all sure were obvious, at least not to me anyway.
Firstly, and yes, obvious now we've seen the uBeam demonstrations, seriously affects usability and user acceptance: the phone needs to be
face down for ceiling-mounted uBeam power transfer to work, so the touch screen and display can't be used while charging. If, instead, the uBeam transmitters are wall mounted, then you'd either need to hold the phone by the edges only, or have it resting on its side on the table. None of these options would be reasonable use cases in my view.
Secondly, I'd not figured out until experimentation that key to the scheme is a sufficiently large area on the handset not only for the energy harvesting to collect enough energy, but also to mitigate localized energy nulls. This part is a non-negotiable part of the design. Until I'd experimented myself, I hadn't realized what importance this had. I don't believe that the energy collection area on the handset can be significantly reduced by technological improvements alone, this is a practical physics problem.