ALL GPS receivers have a hard time getting a lock while in motion, even with a warm-start. This is because it is trying to get accurate timing references from the Satellites (which are moving), while it is moving, and not slaved to a satellite, which means it is using only the internal, unsynced quartz clock. Once it is able to get enough of a reading, it will usually train the internal clock to the satellite's atomic clock. Then it can look at the almanac data to figure out when the other sats are transmitting, and where they should be (ephemeris data). Once that happens, it can figure out where it should be (a second or so after it gets a lock on the first sat), provided it has an up-to-date almanac. If you watch the satellite status screen, you can see the almanac update itself when you see the satellite number jump around.

Now, if a GPSR is switched off and moved to another part of the world, it will need to do all of the above, and it will likely have an out of date almanac. Also, in most cases, the chipset makes a few assumptions about where it thinks it is, so it looks for satellites that the almanac says should be there. When they are not heard, it resets and goes into a cold-start mode. If it is way out, it may take 20-30 minutes for it to get fresh almanac data and gain a fix.

The Wiki explains it better than I can...
http://en.wikipedia.org/wiki/Global_...gation_signals

The space shuttle and ISS use the GPS system to confirm their position, but I don't know if it is primary or not (likely not). There's no reason GPS wouldn't work above the satellites, except that the sats have directional antennas pointing at the Earth.

Oh, and I would imagine that if a commercial airliner is using a GPS receiver, it will be confirmed that it has a position solution prior to pushback!