I think that method of calculation using projection is seriously flawed. By 2040 we are going to see more or less complete switchover to EVs (almost 90% of the vehicles switched to EVs) or very little (2 to 5% of the vehicles being EVs). The second possibility seems less likely with every passing day. I think we are on the verge of major disruption and we are seriously miscalculating EV growth prospectus . Here is why.
Same way when you buy a Tesla model S capable of automated driving, supercharger access and 10 year unlimited mile warranty, you are not just buying a car. You are buying a car (faster and safer) + driver + fuel + maintenance. Now suddenly the cost of model S starts making sense even for upper middle class families.
With this different value proposition, different business models have to evolve to fit the product in your life. May be vehicles for personal ownership will not make sense. May be vehicle as a service will make more sense. May be vehicle will perform multiple roles such as vehicle + power storage. Our thinking and business planning often lags behind technology and that is definitely the case here.
Secondly there are multiple converging trends that are setting up EVs for faster absorption. Solar panel costs dropping down makes electricity cheaper and expands the markets for storage batteries. That trend helps EV and vice versa. Then there is rapid spread of ride sharing services that become more economical with EVs and automation. Then there are unstable oil markets, stricter and stricter fuel standards and pollution headlines. And we have not even talked about the most critical trend, the battery prices. The 200$ / kwh barrier was broken midway in 2016 and if the tech forum chatter is to be believed we are in the ballpark of 150$ / kwh barrier. Beyond a certain point, not only it will make no sense to buy an ICE, but it would also make sense to retrofit your ICE to make an EV. That would burn the ICE candle on both ends. No new vehicles will be ICE and the old ICE vehicles will start being EVs.
This is convergence of several trends that build a positive feedback loop and forms a proverbial perfect storm. This is where conventional wisdom mostly fails. Conventional wisdom is very good at analyzing single dominant trend and making projections. But projecting such convergence is notoriously hard.
And this has happened in the past. As Prof Tony Seba has given example in his video
I think the rapid adoption of EV's is a scenario more probable than we think.
And policymakers and planners should take that into consideration while planning EV future