It would be much more sensible to electrify the cars/busses/trucks/delivery vans that are driven a lot, that are driven in stop-start traffic, and that are driven in heavily populated areas first. That gives you much more bang-per-buck in terms of pollution and CO2 reductions.
Pointing out which specific subcases remain impossible doesn't disprove the larger case.
There's a good question as to how scalable EV's are generally. Present scale is minuscule. A small feaction of a percent in the US, and that actually fell in 2015. I'm not sure what 2016 and future trends (based on pre-orders) look like.
You've got the fundamental problem that battery energy storage densities by weight are ~1/100 - 1/50 that of liquid hydrocarbon fuels. Synthetic analogs of petrol, kerosene, and diesel would be quite useful (though expensive). All but irreplaceable for some uses (heavier-than-air flight, marine propulsion).
Among batteries, you have:
1. Liquid / molton salt batteries. Some of these have highly abundant substrates. The problem is the 300-600C+ temperatures they operate at. Especially in vehicle applications.
2. Metal-air batteries. Iron and aluminium particularly. Here the oxidation is supplied from air. You're consuming the (anode?) in use, and it's got to be replaced, but that can include recycling. Abundant but problematic.
3. Fuel cells. A reaction, typically of hydrogen and oxygen, producing electron flows. The problem is the reaction chamber, which usually requires scarce catalysts (e.g., platinum) which are a) expensive and b) rarer than lithium.
4. Advanced allotropes. Carbon or silicon or other materials which are abundant but offer unique properties in new molecular forms. While this is well outside my area of expertise, it's an un(der) explored area which might pack some suprises.
5. Biocells. Life does some amazing things with enzymes and other agents, including maintaining an exceptionally high voltage potential across cell membranes (see Nick Lane's book, recently mentioned here via Bill Gates). Humans utilising biological mechanisms to provide electricity might offer another out, though this again is highly speculative.
The advantage of my #s 4 & 5 is that they could rely on highly abundant elements arranged in complex molecules to perform desired functions. Experience suggests that such molecules are difficult to come up with, degrade quickly, and have narrow operating bands in which they're viable. But at least the abundance constraint is removed.
There's still interest in this for stationary energy storage on power-grid scale. But not for mobile. Sodium catches fire if exposed to air.
[1] https://en.wikipedia.org/wiki/Sodium%E2%80%93sulfur_battery
Faradion has been working on portable batteries w/ decent energy density (~150 Wh/kg), excellent charge-discharge performance (93% capacity after 1000 cycles) and competitive costs: http://www.faradion.co.uk/about/news/2015/05/489/
Recent formulations for sodium rechargeable batteries are room-temperature and non-flammable: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4525335/
Here's a a short article progress being made getting Na-ion batteries into an 18650 form-factor (their current prototype has an energy density of about 90 Wh/kg): https://news.cnrs.fr/articles/a-battery-revolution-in-motion
While energy density is lower than Li-ion, there's probably a good argument for longevity, safety, and material costs.
Wrt. possible short-term lithium supply issues, i can see how Musk would just build a new mine operation if necessary, like a Giga-ship to mine it from seawater :)
No, it's not. The ultimate causes for costs are human labor and taxes (including similar concepts like rent or license fees) – everything else derives from that.
You're not paying the Sun for its energy output, but the workers that mine the materials, process them, and build solar panels.
Solar flux is a given, and isn't modifiable by human activity. But the flux reaching Earth is limited, and in the sense of that which you give up -- opportunity costs -- there's only so much of it available. Stored solar flux -- as biofuels, wind, hydro, or fossil fuels -- represents an exergic potential (value) with an eMergic cost.
Economic prices fail to consider that emergic cost entirely, but in a very real sense it's an account balance being depleted. Environmental sinks also have costs associated with clearing and negative impacts.
Howard and Eugene Odum, ecologists, coined the term emergy and discuss it. There's a fair literature concerning it in ecology, though little or none in economics. I believe economics is in error here.
Again, very much not mainstrem econ. You'll find treatment in ecological, biophysical, and thermoeconomics however: Georgescu-Roegen, Daly, Hall, and Costanza, among others.
http://www.resilience.org/stories/2013-10-06/energy-ecology-...
(By Howard Odum's daughter.)
But it has to be done only once when the battery is produced, right?
How much is determined by the storage per vehicle, vehicles per capita, lifetime of the battery pack, and recovery rate of recycling.
If you go to an inventory of mineral or element prevalence in the Earth's crust, you'll come up with a list of elements and what percentage (usually by mass) they constitute of the Earth's crust.
Strategic minerals: iron, copper, zinc, silver, gold, tin, lead, mercury, gallium, etc., etc., all have specific rates of occurrence. We mine them from ores in which they're more concentrated, because that's easier (again: less energy requirement), but those ores are scarce. Ultimately the question becomes how much energy is required to access minerals vs. how much energy do they make available. If the first exceeds the second, they're a losing proposition no matter the technology applied to extraction.
Ore concentrations are formed by various mechanisms -- and I understand them only partially, this isn't my field, though I'm studying it now. Iron, for example was largely concentrated into ores billions of years ago, during the first big flourishing of life on Earth, in the Great Rusting. Biological activity, mostly algae, concentrated what had previously been unoxidised molecular iron (sourced from cosmic material that coalesced on Earth).
Other mineral ores seem to also have biological origins in concentration, some various chemical transformations, some geological activity (coal, oil, and natural gas most notably), some are pretty directly remnants of late meteor or asteroid impacts -- especially gold and heavy elements which would otherwise have sunk to the Earth's mantle and core.
An EV essentially is a car with a very expensive fuel tank (the battery), made out of specific materials and properties. While it is efficient in energy use and can make use of fungible electricity, lithium is rare, limited to specific areas of concentration, and can (as any material) only be partially recycled.
There's lithium elsewhere on Earth, or more precisely, in seawater, but processing seawater for minerals (uranium is another which might be extracted) is expensive in the sense of requiring a lot of energy (arguably the ultimate definition of cost is energy requirements).
At US rates of auto ownership scaled to the world and future populations, known lithium reserves would quickly be exhausted -- a few decades, if that. With recycling, yes, some of that would be re-used, but even at 90% material recoverability (and references I have suggest 30% is far more common), you'll lose ~50% of your original material in 7 generations. (The formula: remaining material = (portion recoverable)^generations. So: 0.47 ~= 0.9^7.)
Other options include other forms of energy storage (including possibly synfuels), other battery components, or we all just start walking a lot more.