I once spent most of a 185 bus journey to central London watching traffic in the oncoming lane, trying to work out whether bus lanes or car lanes carried more people.

I think I estimated something like 50 people on a rush hour bus and about 1.25 people per car, then counted average number of cars that passed per bus, and it was somehthing like 25.

So to my surprise, the bus lanes won. The bus lane, despite seeming to be pretty empty most of the time, actually was shifting more people than the congested car lane.