Indeed. It is quite odd (perhaps to the point of being misleading) to use scientific notation ahead of "tonnes of kg", instead of just using scientific notation combined with Kg.

I made this mistake reading the article–I divided 10^7 / 10kg per car --> only enough cobalt for about 10^6 cars (1 000 000) cars.

Instead the calculation is 10^7 tonnes * 10^3 kg/tonne = 10^10 kg / 10kg per car = enough cobalt for 10^9 battery packs = 1 billion cars.

I think we will be ok...