First, there's no reason why advances in manufacturing techniques are not applicable to the J-20. In fact, it'd be expected that it is, as that's the norm and not the exception. As such, I don't see how manufacturing techniques can give the J-35A a significant advantage. Sure, somethings probably can't be changed, but certainly many things can.
I expect J-20A to also have some benefits in manufacturing techniques and maintainability, however as a cleaner sheet design, the J-35/35A may well offer benefits that J-20A cannot take on unless it was more comprehensively structurally redesigned (e.g. additive manufacturing).
Secondly, how many big ticket items have the PLA exported before? The PLA has in the past been constrained in many well known ways when it comes to exporting big ticket items. My suspicion is that it's about to change. The geopolitical situation, product quality, and manufacturing capacity are all conducive to doing so now. The PLA has shown that it can adapt to the times. The first batch of the massive 055 was 8 ships, a break from the previous norm of a few 054s before 054As, or a few each of 052/052B/052Cs before 052Ds.
I think the export potential of the J-35A is an important milestone for the Chinese MIC, and with it there'll be paradigm changes that requires flexibility in thinking rather than just following the old ways.
They've actually offered a fair few big ticket items for export in the past up to now, but usually after they've only entered PLA service.
054A, 052D, J-10C, 039A SSK, even AEW&C -- they've all had domestic equivalents enter service first before being offered for export, and in none of those situations were their export versions greatly impactful on the domestic procurement situation.
This isn't merely a case of following existing trends, but also of common sense -- we know the PLA has been very much "5th gen pilled" since the mid 2010s, and it makes sense that since that time they would have sought the best way to scale up their 5th generation fleet in as practical and timely a manner as possible. As we've observed in the last 5-6 years that has involved increasing J-20/A production scale (it wasn't that long ago that we thought 20-40 per year as a sustainable production rate might be all we'd see), to 100+ aircraft per year, but clearly that wasn't enough for the PLA. If they sought J-35A, we should thus be looking at both the benefits that J-35A bring itself to the PLA from perspectives of cost, capability, time/production -- and "export potential" really should be a secondary if not tertiary consideration, considering how important of a domain (composition/time scale of air fleet procurement) this is.
Simply put -- for the PLA to want to buy J-35A over expanding J-20/A production only requires a handful of things to be true:
- J-35A can be approximately as capable as J-20/A (or even slightly less capable), while offering some benefits in cost reduction (procurement and operating cost, and common supply chain with J-35), AND
- J-35A being able to be delivered to the PLA in a more timely and lower risk way than further scaling up J-20/A production (whether it's expanding CAC production, or asking SAC to produce J-20/A)
If those two criteria can be met, then it's a relative no brainer for the PLA to favour J-35A over further increasing the ceiling of J-20/A production rates