FDMII, re autonomous flight. Capability yes; but cheaper, perhaps not.
Consider the safety cases that would be required for the entire supporting infrastructure; will we be able to afford that, let alone ensure equivalent safety as with aircraft. A CS 25 for every element?
And possibly due to a subconscious regulatory bias, which may give credit for human intervention in unimaginable situations, an even higher level of safety may be required for autonomous vehicles. Compare the ‘accepted’ manual landing risk as indicated by operations (IIRC about 10-6), against the higher requirements for an autoland. Even then there are residual risks with triple or dual-dual systems (pitot icing).
Tourist; always love a challenge:
First separate technology from automation; EGPWS, ACAS, technologies have improved safety.
FADEC, automated engine control. 747-8 and 787 had restrictions for Cb related icing, because the FADEC was unable to manage ice crystals. Worst case was that all engines were affected simultaneously, and with ice damage may not restart. Thus the need to avoid the conditions, not practice flying without power.