At a NASA talk on space debris (part of the Space Technology Grand Challenge program at NASA [1]) the presenter mentioned that so called 'structured orbit' systems (like Iridium and GPS) were particularly vulnerable because their service depended on where they were in orbit and they had finite maneuvering capability. So GPS destruction is a combination of satellites which run out of gas swerving and jinking to avoid junk, and losing individual satellites due to forced (de-orbit) or worse actual destruction.
It was by that reasoning that this presenter was describing the space junk disaster as a sort of slow motion thing but the first indicator that you and I would notice would be that GPS service goes away. Its also exponential in that actual collisions threaten on 1 or 2 additional satellites but dozens in orbits above and below.
Then he finished with current NASA timelines (about 25 years before a system might be tested) and funding (0).
Granted he may have just been trying to rustle up funding for his pet project but I found his reasoning pretty compelling.
If your app survives that long, changing operating systems is a more pressing problem than the immanent destruction of GPS.