Well that might be a bit too 'new' for today's smartphones and rapid app development cycles. However, there is a software known as PTAM that has been around for a while now[0]. It isn't as fancy as the SLAM demo you showed, PTAM can't build continuous 3d maps, but it does the kind of AR stuff we have been promised[1]. It ran in real time on iphones all the way back in 2009!
However, in the real world it might not be reliable enough or might be too costly in terms of battery life to justify putting it in a phone app that was developed as quickly as possible
That SLAM thing is interesting though, especially if you can integrate it into an AR game. Because it can make 3D maps of the environment, you might be able to use players of said AR game to go out and scan the world. One can incentivize scanning unscanned areas by placing 'rare pokemon' in them or giving in game items for areas scanned. If one has enough people playing, one could build google streetview style maps that are updated everyday.
Except we can do much better than streetview because we have the 3d model of stuff. A big continuously updated 3d map like this is valuable. We can use it for mundane stuff like 'does the local bar have any specials today', but we can also use it for stuff like navigating robots. If one has a big map of where all the trees and buildings are, it makes it much easier to do drone delivery.
[0]http://www.robots.ox.ac.uk/~gk/PTAM/
[1]https://www.youtube.com/watch?v=F3s3M0mokNc
[2]https://www.youtube.com/watch?v=pBI5HwitBX4