while seemingly attractive, a 95/96 reconfiguration would involve SIGNIFICANT modification of a battery module, and would likely offset any benefit.
Yeah, definitely not practical.
on the other hand, a 15/16 configuration (ie a 375V pack) is something someone outa work on
Funny enough... "someone" is...
I already have configurations of 13, 14, 15 and 16 modules working in the S+X with the custom BMS I had originally designed for 057's non-OEM pack project. (I actually have a 14 module pack in my X right now for more testing and dev on this hardware... built out of a 2012 85 pack that had two bad modules, migrated to an X-compatible pack chassis.)
This wasn't that hardware's intended use, but it's a functional thing. I've mainly been debating if this is something to focus on making production ready for this purpose, and it seems like it might be time to do so. Unfortunately this currently comes with a lot of caveats, so would take a lot of time to flesh out those details and get it fully on par with Tesla's BMS in such a setup with Tesla modules since this wasn't originally designed with that use in mind.
The biggest problem is there's only so many OEM-valid configurations. Modifying these on Tesla hardware isn't really a viable option, since an OTA updates would basically brick the car at that point, or otherwise be impossible. The custom setup needed has to fake one of these OEM setups,
and be able to handle any updates that come down the pipeline gracefully. There's also the matter of things like supercharging, where the protocol has evolved significantly over the past couple of years and probably will continue to do so (the BMS speaks directly to the supercharger, so the protocol is implemented there), meaning the custom hardware needs a sensible update mechanism as well to be a long term solution.
I've put in significant effort to make this possible, and am pretty proud of the work done on the custom pack project to cover all the edge cases on things like this. Unfortunately that project is pretty much
dead in the water at the moment without someone with pretty deep pockets jumping on board to make it a reality. Migrating a lot of that work and knowledge towards alternative refurbishment options for existing packs is starting to look more reasonable given that the volume of such repairs is steadily increasing and a better solution is very much needed at this point.