Heres a thought:
Bus stop/terminal/depot rechargers+fast charge batteries+highly efficient auxilary power unit=Near electric bus.

Neil -

overhead power lines for trolley buses aren't just eyesores. They also cast the routes in stone. Recuperation efficiency is actually not all that great, unless you have a special grid topology and a bridge to the national grid.

Besides, there is a large up-front investment in the overhead infrastructure that is amortized (eventually) only if the service is accepted. Unfortunately, many US consumers wouldn't be caught dead in a bus - those are strictly for the poor! Light rail systems, especially subways, do not carry this stigma but they are far more expensive to build.

New technologies, incl. HEV/PHEV and dynamic routing, offer possibilities that did not exist until recently. I would be very surprised if trolley buses made a comeback even in places with very poor air quality. On the contrary, if buses are to become more popular they have to appeal to the well-off in terms of aesthetics, technology/green credentials, convenience and comfort. The Dutch Superbus project, for one, is pursuing exactly those aims.

Buses are unlikely to make a major comeback no matter what. The problem is not just aesthetics, it is schedule. They depend on high volume routes for frequent (therefore convenient) service. The newer rendition of the trolly (sorry for the use of that now twisted word), is in dual mode PRT, which potentially can offer the best of all worlds -- private vehicles, door-to door transport, small battery opperated vehicle, and no limits on scheduling.

RUF http://www.ruf.dk/ruf2006.pdf is just one version, but it amounts to a trolly configuration (electric feed at the track) with the flexibility of an auto. The capital expenditure and vehicle adoption are still problems. However, the development of EV's may make the concept acceptable.

That approach has some promise, but depends very heavily on the circumstances. There are plenty of routes here in Boston, for instance, which are only 10 miles long or so, but often times buses do not lay over for long periods, and in many other circumstances, they lay over in locations which are unsuitable for installing such a charge station. A switch out pack that would not need to be changed *every* layover, or one which did not need an entire 25 minute layover to effect a "refueling," could be more suited to such circumstances. Or you could contemplate a more systematic change of the transit infrastructure, which would go beyond the obvious expense of installing a number of heavy-duty charge posts.

Another factor is how efficient is the regenrative braking. 25% is lousy, with an average efficiency of 71% per stage in a 4 stage energy flow setup (1-motor gen mode, 2-battery charge, 3-battery output, 4-motor work output; there are other components, but this is a back of the envelope calc). If the average is bumped up to 91%, then it has a ~68% total efficiency rating. With the first (25%), you can cut fuel expenditure by 1/4 in City driving. With the second (68%), fuel mileage triples.