A couple of comments regarding human and machine pairing.
Firstly, there are six key technological emphasis as part of the road map on unmanned vehicles as outlined by the US Department of Defence (DOD). Autonomy is just one of the six.
There are a bunch of regulatory issues, compliance and technical hurdles to introduce some sort of autonomous vehicle. As an example, there are eight functions required to undertake self separation function.
Whilst there is a roadmap, the road ahead is still very long and winding towards full autonomy.
Secondly, the notion of human machine pairing and the inevitable conversation towards some form of traditional air to air combat as a primary tool is a narrow view of future utility of man/machine matrix. IMO, we need to see such a pairing in the context of a kill chain in the future battlespace. Some of the mission task I see could be :
(a) Extended weapons load carrier
(b) Expanded sensor net in the from of L band frequencies against 5th generation targets
(c) Deep penetrating vehicle too risky even for 5th gen platform
(d) Disruptive asset in electromagnetic spectrum warfare
(e) Conduct deep ISR function
How is this connected to the F-22 program? I think if you look at the key attributes necessary to operate in a dense environment for machine/human pairing , some of the key features needed such as stealth, SA, sensor fusion flight formation and secured LPI communication links are only present in two platforms, the F-22 and F-35.