I see what you mean, but I had thought that computing an average expansion rate would still be analytically useful. Having taken a look at this
I don't believe that's the case. The reason is that it's too sensitive to initial conditions: If I start from 2011 I get an average annual expansion rate of 40% (and that doesn't match the dates between the starting and ending years well at all), while from 2012 the rate is 20% and still doesn't match the discontinuities in the data. That kind of disparity makes this measure useless.
If we compare the F-35 production numbers with where the PLA is now (~30 aircraft per year), China today is where the US was roughly a decade ago. It also illustrates the point you made about experience - the F-35 first flew in 2006 and reached this milestone in 6 years, while the J-20 first flew in 2011 and reached the 30/year milestone (maybe) 10 years later.
Sadly, I'd have to reluctantly concede the point that it's highly unlikely the Chinese fifth gen numbers will catch up to the F-35 - America just had too much of a head-start. But I'm very optimistic about 6th gens as China will start producing those at roughly the same time as the US (certainly with nowhere near the delay from F-22 to J-20) and will have much more wealth and experience as well as a far stronger industrial base than it did when it made the J-20.
The F-35's production scale and the structure of its program is somewhat the exception. The reason why they were able to see a gradual annual increase of production rate year on year, is because of the sheer scale of what their planned full production rate will be (over 160 aircraft a year!)
So I definitely wouldn't use F-35's production scale as if it is a reasonable benchmark for what most fighter production scales should be. Remember F-35 is an aircraft that is expected to have a production run of nearly 2500 aircraft
for the US alone.
In the case of J-20 and J-XY circa 2025 -- well...
1. J-XY is likely to only begin production in 2025 (at the earliest), and regardless of what its planned full production rate will be, it is going to start off fairly low. In 2025 I would be impressed if they produce 20 J-XYs.
2. That means of your 120 5th gen fighters produced in the year 2025, about 100 of them would inevitably have to be J-20s. In which case, you are essentially trying to argue that China can produce 100 J-20s a year by 2025.
As for what a more likely J-20 production scale may be -- here are two projections for production ramp up for F-22, the top table being for a circa 1998 planned total production run of over 330 aircraft, the bottom table for some 176 aircraft after subsequent reviews and cutdowns of total fleet size.
You can see how in both of the tables, from 1998 to the early 2000s, initial production in the first few years was small, before reaching a gradual "steady state peak" of some 36 annual aircraft and 22-24 annual aircraft, respectively, for a few years. Production didn't continue to expand beyond a "steady state peak" indefinitely, simply because the total production of F-22 in both instances wouldn't have necessitated it.
I very much expect J-20's total production to be greater than what F-22 ended up being, and to possibly approach the original fleet size requirements for the ATF program (i.e.: some 700+ aircraft) before it was gradually cut down over successive reviews to the final 180ish total number of F-22s they ended up with.
Assuming a 600-700 strong J-20 production run across all variants, and assuming production would close up shop by the early 2030s, I would tentatively suggest that they could reach a "steady state peak" of 60 aircraft per year for a number of years (say, 4-5 years) in the mid/late 2020s.
Assuming they are able to build 30 a year in 2021, something like this might be plausible (just take it as a general trend, the specific yearly numbers don't matter that much -- e.g.: it might be that the build up between '21 and '26 will be less gradual than that and there will be two years where there is just a sudden significant increase in annual production rate as a result of a new line being commissioned or something, and where the peak sustained rate is 70 instead of 60, etc):
Total cumulative number produced between starting production to mid 2021 -- 50-60 airframes
'21 - 30 airframes
'22 - 35 airframes
'23 - 40 airframes
'24 - 50 airframes
'25 - 55 airframes
'26 - 60 airframes
'27 - 60 airframes
'28 - 60 airframes
'29 - 60 airframes
'30 - 60 airframes
'31 - 55 airframes
'32 - 40 airframes
'33 - end/transition to 6th gen production
The total run here would be 665 airframes for J-20.
Note, the above doesn't include J-XY, which, if it begins production in 2025, depending on how ambitious the PLAN and PLAAF are with it, may or may not achieve a satisfactory production rate by 2030. If it can, then I expect the greatest number of annual total 5th gen deliveries for the PLA may be 2030 and the early 2030s when J-20 production will be at its relative peak and where J-XY production will be ramping up.