I went overboard on that thread with math, so I'll repost it here where it's more appropriate.

My numbers are extremely simplified. I recall an estimate of 350,000 Gen 3 and 150,000 Gen 2. From Gigafactory PDFs, we know that 35 gWh/yr and 500,000 packs are to be dedicated to vehicles at full capacity. Making a huge assumption that all Gen 2 are 85 kWh, and Gen 3 are 60 kWh, we get 500,000 vehicles with total of 33.75 gWh a year. This huge assumption passes a sanity check so far.

Next, from countless threads, it is found that 5 to 10 kWh is not available in the Model S due to antibricking, HVAC, misc, etc. So this means that of a potential 60 kWh battery, maybe 50 or 55 kWh can be used for travel. Using Elon's 200 miles, I get the following numbers.

50 kWh = 50,000 Wh

50,000 Wh / 200 mi = 250 Wh/mi

55 kWh = 55,000 Wh

55,000 Wh / 200 mi = 275 Wh/mi

If the Model S uses an amount of energy per mile that is more than 35% or so than my predicted values at 10 degrees Fahrenheit at highway speeds, my calculations show a larger than 60 kWh battery will be required.

I would then follow up with a 65 kWh Gen 3 battery and see what number of Gen 2 vehicles will be able to have 85 kWh packs, but I would assume most if not all will get the maximum.

65 kWh - 5 kWh = 60 kWh usable (likely too optimistic)

60 KWh = 60,000 Wh

60,000 Wh / 200 mi = 300 Wh/mi

This is easily achieved with a smaller, more efficient vehicle. We're seeing this with Model S under ideal conditions already.

(350,000 × 65 kWh Gen 3) + (100,000 × 85 kWh Gen 2) + (50,000 × 60 kWh Gen 2) = 34.25 gWh

This leaves only 2.14 % of the Gigafactory pack output unaccompanied.

I can't see 1/3 of S and X being the 60 kWh.

Here is another one.

(350,000 × 60 kWh Model 3) + (100,000 × 95 kWh Model X) + (50,000 × 85 kWh Model S) = 34.75 gWh

That's 0.74% unaccounted for.