Flying drones (and thus window-level delivery) would be much easier and more effective to implement for a few reasons: 1) Pathfinding is easier as it doesn't h

Flying drones (and thus window-level delivery) would be much easier and more effective to implement for a few reasons: 1) Pathfinding is easier as it doesn't have to concern itself with roads or obstacles for most of it's route. Just fly up to a reasonable altitude (just enough to clear buildings in the way) and all the drone has to worry about is adjusting for wind and "last-mile" (your window) navigation. Once they start hitting critical mass we'd start seeing infrastructure to keep the drones moving; homing beacons and landing pads on your window, and drone towers (either standalone or grafted onto existing skyscrapers) to refuel(/charge) drones and hoist them up to a high altitude, allowing them to rely more on gliding vs. their own power. 2) Less of a chance to run someone over. Again; since it's not in conflict with existing infrastructure, it's less likely to suddenly find a 6-year-old in front of it. The flip side is that it's going to hurt if the drone fails and drops out of the sky; this can be remedied by a fail-safe (such as a parachute that physically breaks open at a certain velocity; independent of the nav systems) 3) Commercial advantages: besides the general advantages of a drone-based delivery; a flying drone doesn't get stuck in traffic as it's skipping roads entirely. Thus, using a flying drone will give you a significant advantage over your ground-based (be they human or drone) competition re: delivery times. If they aren't faster; they're at least going to be far more predictable.