I'm just providing some constructive criticism here, in the hope this will be used to better run competitions in the future.
The 2017 Visioneering competition seems to have been well thought out and planned with the requirement of a visa. I can see that that was not the case this year though.
More detailed pre-planning could have avoided that confusing situation. Even if the management team hadn't noticed it, someone from the Xprize team internally could have, so please don't blame this on any one actor or team.
Secondly, another major pitfall is that the team failed to anticipate the overwhelming response to this competition, they made the right move in increasing the semi-final winning pool, but yet there still seems to have been a lot of difficulties in picking a winner due to limited winning slots. They could have updated it further by including another quarter-final round prior to the hackathon or even a round 2 hackathon now, to further give deserving prizes a chance.
There might have been designs that would have had a tremendous industry-wide impact if they had been realized judging by the indecision most of the judges felt. A simple search operator on Youtube like so: https://www.youtube.com/results?search_query=xprize%2C+month gives you almost all the video submissions, so naturally would also suggest making videos unlisted.
The really good designs that were filtered out due to lack of winning slots could at least be given a fighting chance with a remote presentation at the hackathon via an on-site representative. This is serious business peeps, we're trying to change the world, most of us just lack the resources to help. Only if this can be catalyzed and the rest of the submissions that almost made it, given a fighting chance in some sort of wildcard round irrespective of challenge area. Xprize would probably have lost the chance to disrupt huge avenues if these prizes had the potential but were just snuffed out.
I hope Xprize takes this feedback seriously and implements it, since the internet is bringing in much more people and connecting us like never before as Peter had stated recently in his blog, we're inching closer to the steeper sections of the exponential gradually! It's time to scale up! Best wishes!