> Public transit depends on the assumption that some trips are more common than others.
Public transit depends on the assumption that there is enough density along a given route to justify its existence. Take a look at the NYC subway map. In the highest density boroughs (Manhattan and Brooklyn) the routes basically go from everywhere to everywhere. Even more so for the Manhattan bus map. That's what you want in a large dense city.
In smaller cities, the "build a high density core surrounded by lower density areas" model is the thing that causes more congestion, because then the core ends up as a bottleneck but people don't want to take transit to get there because it doesn't have frequent service to the areas outside the core at one of the traveler's endpoints. For those cities it's better to have medium density everywhere than try to make transit work in a city where a large proportion of the population is coming from an area with density too low to make it viable.
And if the whole "city" is low density, i.e. it's a rural small town, then it's not likely you're going to make public transit work there whatsoever. The best option there is to use mixed zoning so people so inclined can live within walking distance of shops.
> It's not a Pareto improvement, as the traffic will become worse for those who drive on that route. But it's not a huge loss for them either.
It is though? The premise to begin with is that road is already too congested and is slowing down the buses. Removing a third to half of its capacity is going to make it dramatically worse. That's what many of the proponents of bus lanes are after -- they want to force people onto the bus by snarling the cars.
They refer to this as "induced demand" by inverting the sign when what they really mean is to suppress demand for driving by making it more miserable, but don't want to call it that because it would be unpopular.