I do agree with the intention ,and we will see small iteration and small production of more advance models.
And with the expansion of the Bohai production facility is for PLAN to build better bigger and mostly more SSNs.
But my argument is , that the time table of which you are comparing to the advancement of surface fleet might not be comparable .
And as you put it the "pulling the trigger" on next generation SSN and SSBN won't be in this decade.
Reason being , nuclear reactors ,nuclear power station has only matured this decade for China.
To make it much much smaller , quieter is a no small task , and that would take time.
To make a SSN you needed a fully matured nuclear industry .There is no skipping this step in my opinion , for example UK was helped US during in the 1950s to speed up it's own reactor development.
The same goes for the next generation ballistic missiles.
The same hold true for the surface fleet ,during the 1990s Chinese electronics manufacturing matured, and gas turbine was bought.
Only as recent in 2010s Chinese surface fleet have been fitted with Chinese turbines. But they have been using Chinese radars since the 2000s cause how Chinese quickly fitted itself in the Chip fabrication chain.
I think the Chinese SSN development using your analog is somewhere in between 052C to 052D right now ,meaning next generation will be 052D , a competent platform that the PLAN is willing to make more then a dozen.
I have no doubts that these hurdles can be overcame with time and resources, but since the pieces that China needed for the development of the next generation only came to be in the 2010s , I would think that the next generation will be ready in the 2030s.
I generally agree with your arguments, except I think that my expectation for "the next generation" is perhaps up to 10 years ahead of yours, because I believe the relevant major industries and subsystems for a competitive SSN/SSBN capability (including modern nuclear propulsion for submarines) would have been ready sometime in the 2015-2020 period -- i.e.: nearly three decades between when the propulsion and major subsystems for the original 09III would have first been developed, using technologies and industry of China in that era back in the early 1990s.
I think 09IIIB could be something like 052C to 052D, and that shortly after we will see 09V that will be something like 052D to 055 (though as I had written to others before, the analogy obviously isn't perfect).
If your skepticism surrounding the idea of 09V emerging in the 2020s is because of immaturity of nuclear propulsion or other nuclear submarine specific subsystems, keep in mind that it is nigh impossible for us to visually identify advancements like we can for other systems such as destroyers, so the best that we have is correlating tidbits of official information with credible rumours and real world documentable progress, and with some general common sense.