@StopcrazyPP I agree and realise the differences (though in fairness didn't know they'd gone to LG, thanks for that bit of info).
Real back of an envelope stuff, but let's say it's two weeks to build a 3.0 pack, two guys working on it. Providing they have enough order coverage to do 26 packs a year. Employ 2 guys @ $100k per year and that puts each pack at around $7k in labor. This leaves $22k for cell purchase.
So even if they are paying half for the Panasonic S batteries compared to the LG chem, there is still no margin in it for Tesla.
Now of course this is full of assumptions, there is always the possibility that the 3.0 batteries are priced so high because actually Tesla don't want to do the work, and when they say they are doing it "at cost" they have worked out the labor cost number on the basis of getting JB to wire the packs up himself
This $12k figure is really just a hangover from the Roadster launch, where Tesla pushed forward a figure that aided in the theory that battery costs would have declined to this sort of level by now, as well as remove a valid concern of owners that continued service would be well supported.
Now the GF might make the former happen, but I'm still not clear how it will help in ensure ample supply of cells using today's battery tech in 8 years time, as surely Tesla will have moved onto better chemistry by then.
The truth is I fully expect the S packs to outlast 8 years, but effectively be a lifetime item of the car. Ultimately if a pack costs more than the value of the car (which will likely happen for 8 year old S's), then it becomes somewhat moot.