In the first level analysis.

Once you realize the sheer amount of commerce and business that is enabled by roads you see that they pay for themselves several times over.

Once you realize the sheer amount of commerce and business that is enabled by a decent metro system you see that they pay for itself several times over.

That's even easier to measure because these gains are localized.

Right. So let's have both! You can't replace the capabilities of roads with metro and you can't replace the convenience of a metro with roads.

Specifically, though, viewing roads or metro as a "loss leader" is probably an inappropriate analysis.

Actually, it has been shown that removing cars increase commerce.

This is a universal outcome with research to back it up?

yes but I am about to go to bed so the link mihht come only tomorrow.

Also some studies have calculated the impact on health in the society and calculated a cost to the society per km of driving a car and the money saved when doing the same km walking or with a bicycle.

The difference is that the gas tax or similar user fees could easily (if politically feasible) be scaled up to cover both capital and operating expenses for the entire road and highway network.

If you tried to do that for merely the operating costs of many transit systems, they’d enter a death spiral.

From what I'd read, the majority of road wear comes from tractor trailers, who don't nearly cover their costs.

It may help to either tax them directly, or indirectly via diesel. Yes, costs pass to consumers, but it would also encourage more done via ship and train, I feel. Even that would be a huge help to clearing up traffic and lowering infra spend.