This sounds similar to what I have heard with regards to the semiconductor design industry. The initial basic design is done with an extremely small team of people (2-3 people max) for like six months or a year. Then that team gets increased to like one dozen or two dozen people who work 2-3 years to mature the design concept, only later do they bring the whole design teams to work on details which might be a hundred people or more. Typically once a design is finished on the mature design stage then another small team is selected to begin to design the next generation. Smaller organizations only have one flow line like this. But larger organizations have like 2-3 basic design teams to weed out bad concept designs at an early stage. They also might have a separate team which is kept working on refining an already produced design while the new design still doesn't require all that many people in it and is far away from production. Those people are then moved to the new design team once it is mature. This ensures you have at least one competitive design always in production.
Since China follows the Soviet practice of having multiple design bureaus competing against each other at the design and prototype stage the 2-3 internal basic design team approach is IMHO not useful. In this case you have CAC and Shenyang for fighter design. Something similar applies to the trainer aircraft where you have two companies competing. However in the case of the large airplane design teams, like Xian, I don't know how they operate but that probably is a limitation. In that AFAIK Xian have like no competition in the system. So the only way to inject competition is to buy a design from outside, as was done with the Y-20, so they can compare that against their own design, and make improvements. The Soviet Tu-160 for example, was a result of a competition where three design teams proposed a design: Tupolev, Myasishchev, and Sukhoi. In the end it was decided the Myasishchev design was superior but because Myasishchev had no experience with detailed design of large aircraft Tupolev was entrusted to bring the Myasishchev design to production. Which they did and it was a successful aircraft.
That is a weakness in large Chinese aircraft design IMHO. Either Xian incorporates more internal competition into design, or China will have to continue importing designs from abroad, I also think the window might be closing on China relying on Ukraine for technical help on this regard since the USA is increasingly asserting its power there. Ukraine is presently under a huge IMF loan which definitively place them under USA vassalage. I think the fact that China did not import Ukrainian aviation turbine technology when they could and overly relied on their native design was a missed opportunity for example.
Progress and Lotarev were the leading civilian engine design companies in the Soviet Union. We also heard a lot about licensed An-225 production, for example, but I assume that was a bit of a red herring since An-124 production would be much more in China's interest and both use similar technology.
Anyways, assuming the practice is similar to the chip design practice I said above, I assume there are at least two teams at CAC right now. One working on J-20 design improvements, as was done to the J-10 in its many iterations, and another working on a new design, which I assume can either be the FC-31 competitor aircraft or a novel 5th "Chinese" generation design. Assuming there are new J-20 revisions, which features do you think will be most likely? Besides the new engines. I would assume either an improved radar, or improved situational awareness either with helmet mounted sights, automated AI pilot assists, or extended range versions. I could see it in an improved version supplementing the J-16 in the fighter bomber role for example.
