Sure there is some lag introduced if it is remotely operated, but then it has the benefit that without a pilot on the inside, it can preform maneuvers that would cause a human to black out. Plus you can have an unlimited number of copilots looking at radar/video looking out for other aircraft.
I don't believe that an autonomous computer program today is capable of deciding when to shoot missiles at people, when to shoot down another aircraft, or when to drop bombs on a building. I don't think a program, with no human input, no GPS, at night, in inclement weather, in the face of enemy sensor jamming, is even able to decide if it's reached its target accurately enough to then attack that target. I also don't think any computer program, even operating only in a simulation, can perfectly evaluate the acceptable level of collateral damage to civilians and infrastructure from its decisions -- to decide, for example, whether it should complete its mission of bombing an enemy unit if that unit enters a home or a hospital. I also don't think either of us is more informed and smarter than the US military and US congress, both of which have decided that there is still a future for manned aircraft for all these reasons.
I don't believe you really think otherwise either. "At this point", we don't even have autonomous cars that work in the real world. With all the best minds in AI working on the problem, with a working GPS and a paved road, our "Google cars" are still foiled by such simple things as a foggy morning blinding the LIDAR. It's going to be a long time before you replace the human pilot in an F15 or a B2.
Yes. Drones and long-range weapons delivery (either high energy or SCRAM-driven missiles). Fighter jets? Might as well call out the buggy whips and horse drawn carriages.
Obligatory A-10 shoutout: We still need something that can provide air-to-ground support to ground troops with a long dwell time, but high-speed cruise to fast deployment (shamelessly stolen from The Avengers: http://www.impdb.org/images/a/ae/Avengers2012CGI_2.jpg).
Of course, designing the same airframe to replace both the Harrier and the F-16 is a ridiculous notion and the F-35 suffers for it.
Frankly, for the type of wars we're likely to fight in the foreseeable future, the most common mission would call for something like a cheap prop-driven plane with a ton of ordnance hardpoints. Take the flight surfaces off, stuff 3-4 of them into a C-17 and fly them to the theater of operations, go to work.
But the aviator community would hate it.
The role you're describing is currently filled - very well, I might add, by the A-10. It can fly low and slow (has a stall speed of right around 140mph), it's heavily armored and can carry massive amounts of armament. Not to mention an enormous gun that can decimate pretty much anything in its path.
They're slated for retirement soon, though. Long term, I think this role will eventually be filled by drones.
I was expecting someone to jump in with the dog-fight argument against using low & slow aircraft. Unlike in Vietnam, today's missiles are reliable and effective, so even a prop plane can engage and defeat a jet at extended ranges... as long as positive hostile ID can be made early enough.