Here's one study, http://www.environment.ucla.edu/media/files/BatteryElectricV... - look at Figure 1 and 2, battery & engine so dominate the manufacturing footprints that everything fades into insignificance.
Of course, if you make the car really light you can somewhat reduce the required engine & battery. But not drastically as weight has little impact at cruising speed & you have regenerative braking in city.
That makes no difference because cost is a relative measure, not an absolute one. If you priced in CO2 then everything would be more expensive, but the relative difference between items would be unchanged. (And that money would go back to the citizens giving them more money to buy stuff, so effectively you did nothing at all. It would only work if there were low CO2 alternatives, but there are not any. All the renewable energy sources have decade long payouts, so at least initially emit lots of CO2, which you would have to pay tax on, so the tax makes renewable energy more expensive too.)
Because ultimately everything requires energy to make cost is an excellent proxy for energy, and in a world where most energy is from hydrocarbons, cost is a proxy for CO2.
Which one is accurate then ?
Note that the slide is not talking about the same measure (co2), the slide screenshot is titled "total environmental system damage" which smells a bit like picking an rarely-used but favourable metric here.