The key point that is likely tip the scale in one direction or another is the positive externalities of converting a driver to a rider. This externality will differ by city, line, time of day, etc, and it would include external costs of traffic, parking, car accidents, among others. Economists can approximate this sort of thing. We are often really bad at accounting for externalities because they aren't immediate dollars in or out of our pockets and lack certainty (pollution controls, infrastructure improvements, etc), but this case may be a bit more straight forward.