> If you could move an entire country from being car dependent to being largely car free, by pushing rapid urbanisation and transit and you put 20 car drivers onto a train or a bus, then we're starting to talk about measures that actually have some effect.

This isn't necessarily true. Depending on how you put "20 drivers onto a bus" in a dense rapid urbanization, you are often generating more CO2 than the drivers were previously. Building an entire nations worth of new train lines generates a lot of extra CO2. "Rapidly urbanizing" everyone generates a ton of extra CO2 (and will instantly price out nearly all of the residents you want to 'urbanize', something EVs never do)

In the US, most EV drivers with a renewable energy source are already emitting less CO2 per trip than the equivalent bus ride would emit (even after factoring in per-rider numbers for the bus). Depending on your area, the "largely car free" option is semi-regularly worse for the environment once an EV is factored in.

> Or if we put some significant money into a manhattan project for carbon capture, that at least gives us a statistical shot.

We already are? https://carbonengineering.com/ and https://www.climeworks.com/

That doesn't discount the work people are doing on EVs. Its still usually better to prevent emissions than to attempt to recapture them later.