I agree there is going to be a huge demand for EVs, but Tesla is already supply constrained... that is why it's more important to look at the time it takes to ramp up battery and mining supply than looking at it from a demand perspective. The issues Tesla is having with Panasonic not supplying enough batteries, and Elon talking about getting into mining are hints that these are the things which are stopping Tesla from expanding faster. And this is with EV's only taking a few percent of the production share.
When Tesla decided to build the Gigafactory it was clear that without it Tesla could not succeed because there were simply not enough batteries being produced. So they spent 5 years ramping up the Gigafactory.
Now imagine how much more supply constrained the raw materials for batteries are going to become over the next 10 years as EVs start to become mainstream. Tesla is about to face a similar issue with the supply of raw materials as it did for battery supply. There is simply not enough global capacity for both batteries and the raw materials for batteries to support the rate at which EVs will want to grow.
I want EV's to take over the world as much as anyone, but raw materials and battery supply are on the critical path and will take a long time to ramp up (especially over the next 10 years as we enter the steep part of the S curve)
I still think demand is by far the most important consideration. 100% of demand will be for EVs once they are cheaper upfront, in addition to being much cheaper to fuel and maintain, with much longer lifespan (hence lower depreciation) and much higher safety (in addition to not killing millions of people per year from air pollution and contributing to global warming). This threshold has been passed already for sedans priced $40k+, but it should be passed for all vehicles in all regions somewhere between 2023 and 2026 driven by battery and EV powertrain experience cost curves.
As you say, this still leaves a lag before supply can catch up. In this period some consumers will buy ICE cars because all EVs are sold out, while some will defer car purchases causing a decline in overall car sales.
Typically design to mass manufacturing took 5-6 years in the auto industry, but Tesla achieved c.3 years with Model 3. Hopefully it is going to prove with GF3 in China is can go from swamp to mass manufacturing within one year when building an existing model in a new location. Other auto companies are still much slower and resistant to EV commitment, but it should still take less than 5 years to switch all production to EVs once EVs are profitable and once consumers demand them.
Cell supply and raw material supply ramp are potentially the slowest to achieve, however the cell megafactory pipeline for 2028 is already at over 2000gWh and rising rapidly. And hopefully Tesla will again massively speed up the cell ramp process when they take cell production in house in the near future.
Lithium carbonate/hydroxide and EV grade Nickel mines/processing plants are lagging currently. A lithium plant normally takes 6 years+ to get to production, but there are plenty of resources available, all that is missing is capex in mines and plants. If lithium or EV grade Nickel become the only bottleneck to a $3trn+ industry there will rapidly be a huge amount of R&D and capex in building these facilities and researching new methods to produce more cheaply and ramp more quickly.
Probably around $600-800bn capex will be needed to build the battery metal plants, cell, pack, powertrain & car factories from scratch needed for 50 million annual EV production (likely enough for 90%+ of the car market given osborning). This can easily be funded by Tesla, EV startups, battery companies, capital markets, Apple, Amazon etc alone if the current ICE automakers continue to resist, but I'm sure the surviving ICE OEMs will cave in to massive EV investment too eventually.