> That makes sense - the limits are designed to stop you guiding a missile with one, not guiding a jet plane.
Only some kinds of missiles, though. It seems like those limits would be quite acceptable for cruise missile guidance, and I'd think GPS would be more useful there than in an ICBM.
I just had a very dark thought suppose a nefarious entity got hold of starship, added a multi ton nuclear payload and sent it towards some city center.
How does it get prevented?
It's much easier to intercept the missile earlier rather than later, even if the window allows for a later intercept.
It's easiest during ascent phase, because: 1) the rocket is still accelerating (slower target), 2) the rocket is still burning (bright target) 3) before the payload(s) has separated from the booster (big target) 4) before warheads and decoys deploy (single target)
There are some other aspects to consider as well, like a trajectory-changing near-miss during early ascent will have a much greater impact than the same in late ascent, (depending a lot on the target type).
Which is why US missile defense began as an entirely ship based program, to be deployed where ascent interception would be possible.
Although this has changed a great deal since then; technical abilities and limits revealed through testing, as well as political/ budget constraints have shifted the tactical role of the system more state-side, not without controversy.
And no, cruise missiles are not scarier than ballistic missiles. There is no difficulty detecting them or distinguishing them from other aircraft. Ballistic missiles are far harder to deal with because the time to intercept is tiny and they have gravity on their side.
If you have the capability to acquire/build a ballistic missile of that class, you most likely can (will) build your own GPS/GNSS receiver. Hell you can take a software defined radio and make your own software-based GPS receiver that does NOT disable itself at those described thresholds (some folks have done this for extremely high-altitude balloon experiments).
Selling it of course is another question.
Keep in mind, the programs that evolved into NAVSTAR started in the 50s\60s and the tech specs were pretty much set by 1980, along with the legal framework and military policy regarding it's operation.
Ironically, it was some of the systematic weaknesses if the original design that led to it being so much more useful (and accurate) than it was ever intended to be.
But it is still a military program, with military style regulations, so it shouldn't be too surprised that some of the 1960's era rules haven't evolved too much.
It's more surprising to me that so many of the restrictions have been relaxed. (Even if it's policy playing catch-up practice.)
It's not Joe Average will try to make a missile with a GPS from his Garmin watch...
Obviously drone attacks make use of GPS, and it's a worry to many -- for example, there is frequently GPS jamming when Putin is visiting somewhere.