I think this is a great proposal. Even so, I'm curious why there needs to be a formal process for the use of many sources in localization. It seems to me that anyone making a localization device (i.e. Apple or Samsung making phones) could add inertial information, cell tower locations, and more to give great location info.