Right now, in in the foreseeable future, no UAV can outperform a jet flown by a fighter-pilot.
Besides, there are many external environmental factors that the pilot experiences when engaged that help him max perform the platform, not the least of which is 360 deg situational awareness, airspeed (energy level), G-loading, weapon envelopes, thinking ahead from where you are to where you are going (or want to be), all these come together to give a complete picture.
Until UAVs can provide that seat-of-the-pants total environment feel and sensory input, there is no way a UAV can even come close to a manned fighter.
How about we field an unmanned tank on the battlefield, first, after all, it is a 2-G, 30-kt, two-dimensional maneuvering platform so it should be easier.
Learn to walk first, then fly.
I think drones in the form of really, really smart “missles” with a primary turbine are going to be a loiter threat in the very near future.
Autonomous operation is reaching exceptional levels that exceed human capability. In my opinion it is the software that will advance the capability. Simply retrofit existing “hardware” platforms or design new higher performance ones. The concept of swarms has significant advantages over a single high performance aircraft even if individual swarm units do not perform at the same level as their targets.
Give it ten years, these platforms can be put together much faster and cheaper because they don’t have to account for human limitations or safety concerns (and the regs that go with it).
http://edutube.org/en/video/autonomous-quadrotor-helicopter-aggressive-maneuvers