You’re still going to wind up assuming a lot of baseline common knowledge that, while it might be stuff you find interesting and have come across in pursuit of your personal brand of geekery, are likely not actually necessary for the job.

Just consider that there are people - perfectly intelligent, capable of learning - who don’t actually know what ‘GPS’ stands for. They maybe don’t know that the GPS system is separate from their phone’s cell connection. They may not even know that satellites are involved. They may have not taken enough of an interest in space technology to have internalized how satellite orbits work, to have intuitions for how orbital speeds and altitudes are correlated. Or they may have picked up a common misconception at some point in the past about how cellphones work, thinking cellphones are always talking to satellites. So they might have made some reasonable intuitions about how GPS works that - because they haven’t been exposed to the true answer - cause them to make some erroneous assumptions that seem like dumb mistakes a poor engineer would make.

Not having had the opportunity to come across those things is not the same thing as not being able to incorporate that knowledge into your worldview when you encounter it.

Ok, but feeding the candidate half of the information is pretty different from coming up with the rough design with “no idea” how it works. Yes, if you basically tell them how it should work, then they can probably figure out how to calculate the time difference and do some trig.

I mean if the candidate would insist that the smartphone sends a signal to the satellite, you can go on and ask "what kind of data?" -"Nothing, just a PING to which the satellite answers with a PONG that contains its position". This is not at all how GPS works, but this system would also work (in a sense you could determine your position on earth based on the RTT + transmitted satellite coordinate). It is not that the candidate should come up with a perfectly correct solution (actually this could be bad because he could just blindly repeat what he/she read without understanding it) - its whether he/she can come up with a solution for the problem at all. In the interview you could still go into discussion about that approach then (what about undirected energy dissipation for contacting a thousands-of-km away satellite, and problems with time-multiplexing transmissions now on the satellite that has to answer individual requests vs broadcast etc.).