Samuel Arbesman (who may have a tad too much time on his hands) has developed a transit system for the Milky Way:
It's a cute idea, but I have some issues. The biggest is differential rotation. In much of the Milky Way, the tangential velocity is constant (i.e. in km/s). However, this means that the inner parts of the galaxy are rotating faster (in degrees per second), such that the whole transit system will be winding up, making more and more unnecessary trips around the center in the inner regions -- this is the classic "winding problem" of spiral arm theory. If you fight this by making your transit lines fixed, then they'll go out of phase with the spiral arms, leaving a transit system that serves the thinly populated rural regions of the galaxy, with no regular service to the dense urban areas. Finally, there are useful, well-populated regions above and below the plane of the Milky Way disk, so there needs to be a three dimensional component to the transit systems, definitely requiring some sort of bus-rail-submarine link plan. But, we've got a infrastructure-based stimulus package coming up, so maybe we'll get lucky.
