It's one thing for a single model to gain 10% of a particular narrowly defined segment, quite another thing to capture 10% of the entire automotive market.
I certainly agree that there is a lot of emerging demand in developing markets. Will this rising middle class all want to drive the same car? Perhaps, but 10 million seems a stretch to me. If the Model 3 were priced 50% lower, your suggestion would be more compelling. The $35k to $70k price range is pretty high end for an aspiring middle class in the developing world. As a fleet vehicle, however, you could get enough fuel displacement value out of a long life battery to make this work, but that is the sort of thing you would find in a purpose built commercial vehicle. I don't think Model 3 is sporting a million mile battery at this point in time (not to say it couldn't at some point in the future). I'd also point out that in countries were the pay scale is a tenth of what it is in the US, FSD is not nearly as valuable because human drivers are not expensive to hire.
So sure, while I am optimistic about these opportunities and think that the Model 3 can reach 1M to 2M, I struggle to see how this scales to 10M.
On the supply side, 10M Model 3 would require some 750 GWh of pack supply. Just scaling up to that sort of volume would take 6 or more years. But hopefully Tesla will bring other product to market in that time rapidly increasing demand for pack supply. Let's suppose Tesla can get to 1000 GWh by 2025. Do we really want Model 3 to consume over 70% of that supply? That would really not leave much capacity for Model Y, Tesla Semi, Power products, or anything else. I think Tesla may do better with a more diversified product mix, which also will go a long way to tapping these growth markets in emerging economies. I think aggregate demand for battery pack grows faster with a diverse portfolio of products than with a single killer product.