We all need to put this in perspective. On battery day Tesla explicitly stated they would use different chemistries for given purposes. There is every reason to suggest they'll try to evolve to a standard format that can be deployed for all chemistries. We might think of that more akin to the commercial paint suppliers that have one to several bases, and supply color on a customized basis. Tesla could very well follow the CATL practice which does cell to pack on a generally standard basis, but can fill the cells on a customized basis.
Tesla already has done that, but ended out with a need to be flexible due to acute cell shortages. With Panasonic and LG both working to supply 4680's and CATL supplying elegant cell to pack designs it si likely that Tesla structural packs will end out vastly cheaper, lighter and more efficient, not to mention being more flexible.
Unlike some of us, I'd happily wager that by 2024 Tesla will be pretty much standardized in form, but with substantial differences in chemistry and cell design from application to application. That will allow standardized and highly scalable production lines for vehicles and stationary packs, while having optimized cells for specific applications. Further, I strongly believe that from Model '2' to Model X, Cybertruck to semi, Powerwall to Megapack, the fundamental pack designs will be standardized. This will allow rapid changes and new deployments coupled with much simplified scale.
During the Musk-Diess conversations this topic must have been featured. VAG has been highly successful with standardizing of many components from Bugatti to Fox. Dimensions and modularity have been huge advantages for them. Tesla has done that with motors from SpaceX to Model 3. The problem has been scaling of battery suppliers
Those all solve within three years at worst, more like two years most likely.