Look at the requirements for each type of mission.so let me sum up, what you said is that UCAV is going to be so advanced in the next 15 years and been able to replace the manned fighter?
Well, yes and no. Basically I agree there might be 600+ J-20 in the future. but been working with machine learning or so-called "AI", for a few years, it's still very difficult for me to imagine UAV could completely replace manned jet in only 15 years. At best it will be like many J-20s/F-35s/Su-57s or even large stealth aircraft (with the shape like B2) fly alone with a lot of drones, and that's exactly what China/US/Russia is developing at the moment. Thus for me, a stealth fighter will still be able to play a pretty crucial role as a data center and decision-maker in the front line network, therefore as long as Chinese want to extend their reach beyond 1-2 island chain (which I personally believe they definitely wish to and have already been working so hard on it), they need a J-XY.
By saying 20/50 years, I was only trying to emphasis the PLAN's ambition for a blue water navy. So if they could operate that an aircraft carrier group far away from the mainland in the 10-15 years, I think J-XY is a very economical cost for them to pay.
Also, it gives me the impression that you think China can't develop a 5th gen carrier-borne fighter and other types of equipment like a next-gen fighter at the same time? Well, let's just say this, it's almost 2020, not 90s anymore, like the US, China is now able to R&D so many toys at the same time without picking an "important priorities". Probably not as hyper as YF-22 vs YF-23 or X-35 vs X-32, but just recall what they show off in the UAV thread recently, it is pretty impressive that so many institutions have the resources to present their own work.
For land attack, surveillance and air to air missions, I definitely think UCAVs will be better than manned fighters.
But for antiship missions and overall C&C, it may be better to have a human in the loop. But that may not be the case.
So the best solution may be drone fighters/bombers with overall direction from an AWACs or manned fighter.
You don't need that many manned aircraft then, say
China can develop the JH-Y, but even a J-20 derivative would still cost a bare minimum of $2B. But is it worth it?
It's good to see so many concepts being researched, because the competition forces everyone to work harder to justify their proposals.