This is fixating on where the ball was, not where the ball could be going.
Design shifted away from maneuverability in the 90s because upper bound of manned maneuverbility was reached. Answer was to increase capability gap was pivot to stealth, sensors fusion, ew, bvr etc. When unmanned performance envelope opens up due to beyond man tier AI piloting, there's chance design will circle back to exploiting new kinematic options in conjunction with other 5G/6G advancements. It can be both, not either or. The energy math for defeating modern missiles is to outmaneuver them from afar thanks to sensors so they never reach no escape/terminal engagement range, which will lead to increase in missile complexity, which is fine, since it raises technical floor/expense of engaging next gen+ hardware and reduce less pool of capable adversaries. For me, this is possible direction for _POST_ NGAD projects. Keep in mind NGAD is mid 10s "legacy" thinking, along with associated visions of CCA/MUM-T. It wasn't concieved based on AI capabilities we now have. The point of NGAD is manned penetrating counter-air ... there's no reason to believe after NGAD won't be unmanned penetrating counter-air, and if that's the future, good reason to dump points in kinematics.
Hence I would not be surprised if future penetrating unmanned platforms = more performant airframes because they can cut down in other components, i.e. disposable, high performance, attritable shooters supported by sensors in the back meshed by flexible datalink. Especially considering wear/maintenance curve will be different for unmanned platforms that doesn't need regular flying hours. No need to support expensive engines and airframes with 8000 hour lifespan when you can cheap out on 500 because you don't expect platforms to survive more than a few sorties against peer adversaries. There's places to cut costs when designing for high performance attrition, i.e. simply having moderately sized airframes go on one way missions solves a lot of tanking problems in IndoPac.
Design shifted away from maneuverability in the 90s because upper bound of manned maneuverbility was reached. Answer was to increase capability gap was pivot to stealth, sensors fusion, ew, bvr etc. When unmanned performance envelope opens up due to beyond man tier AI piloting, there's chance design will circle back to exploiting new kinematic options in conjunction with other 5G/6G advancements. It can be both, not either or. The energy math for defeating modern missiles is to outmaneuver them from afar thanks to sensors so they never reach no escape/terminal engagement range, which will lead to increase in missile complexity, which is fine, since it raises technical floor/expense of engaging next gen+ hardware and reduce less pool of capable adversaries. For me, this is possible direction for _POST_ NGAD projects. Keep in mind NGAD is mid 10s "legacy" thinking, along with associated visions of CCA/MUM-T. It wasn't concieved based on AI capabilities we now have. The point of NGAD is manned penetrating counter-air ... there's no reason to believe after NGAD won't be unmanned penetrating counter-air, and if that's the future, good reason to dump points in kinematics.
Hence I would not be surprised if future penetrating unmanned platforms = more performant airframes because they can cut down in other components, i.e. disposable, high performance, attritable shooters supported by sensors in the back meshed by flexible datalink. Especially considering wear/maintenance curve will be different for unmanned platforms that doesn't need regular flying hours. No need to support expensive engines and airframes with 8000 hour lifespan when you can cheap out on 500 because you don't expect platforms to survive more than a few sorties against peer adversaries. There's places to cut costs when designing for high performance attrition, i.e. simply having moderately sized airframes go on one way missions solves a lot of tanking problems in IndoPac.