Both autonomous cars and autonomous aircraft (likely) have manual backups. So the goal is to be 99% autonomous, or something like that. That works for flying because the failures are probably more far between. The planes are very well maintained, the route is clear from other planes etc. For a car, reaching 90, 95 and 99% is really difficult because of other drivers, bad weather, ice covered sensors, higher risk of mechanical failures. I agree making a 100% autonomous plane that needs to land safely in an emergency is probably more difficult for the reason you mention.

>the driver can just pull over and stop

Yes, sometimes you can pull over and sometimes you just have to put the blinkers on, stop in the traffic lane, and hope for the best--even though it can be very dangerous. One challenge is that, with full autonomy, the vehicle needs to be able to very quickly deal with a once-in-10-years scenario in the least bad way--both for its occupants and for any other traffic on the road.

Sure, thing is these issues are extremely rare (and not just over populated areas) And that is why you still have pilots, and redundant systems, so yes as a system an aircraft is a lot more complex. The traffic environment is not, the high level of control of air traffic and its standardization is making sure of that. And the latter is making i easier for aircraft because they have to deal with a lot less randomness.

So, I could imagine developers of self-driving cars could learn a thing or two from aerospace. But that is just my opinion.