I often hear that Human pilots are superior to machines. Apparently there are thousands of Engine failures caused by ingesting birds; since I know birds can be detected by radar hundreds of feet away, it would be possible for an autopilot to avoid birds - while it is hugely difficult for pilots to "see and avoid" birds.
This is to say nothing about bringing disabled airplanes back to their airport safely rather than resorting to a statistically fatal water landing.
One of the features of a peer to peer autopilot/autoATC is the ability to 1. reroute multiple planes to avoid enemy aircraft (ie birds) in real time, and 2. reroute multiple planes to prioritize the landing of distressed planes.
It is my opinion that on second analysis, this lucky landing will be criticised as a series of poor choices, and point to 1. the inability of the tower to route traffic around a flock it should have seen, and 2. the inability to react within the time demands of a predictable event. (Since the climbout for this plane is higher than the glideslope, the pilots best choice was return to origin; he appears to have had about twice as much speed and altitude as he needed to land at la Guardia.)

Are avoiding birds, and prioritizing dead stick landings strong arguments for increasing the role of machines in the ATC?