I would like to know the feasibility of an interplanetary "bus" or "taxi" which moves between, say, Earth-Mars or Earth-Venus without entering the orbit of either. The idea is that the taxi is large--more like a roving space station. It would have thick, heavy shielding to reduce radiation exposure to Earth-surface levels, enough interior volume to maintain full greenhouses to feed, say, dozens of people indefinitely, and full life support systems for recycling water, CO2, etc.
Travelers would need a conventional rocket/spacecraft to rendezvous and dock with the taxi, but the launch/land vehicle only needs enough life support to meet up with the taxi and then hop off at the destination. Clearly, the taxi would need to move at a velocity within the delta-v budget of the launch vehicles, even though it could theoretically take an arbitrarily large time to accelerate to any desired velocity.
I realize that such a taxi would be immensely expensive and difficult to build. I'm not really concerned with the economics. I'm more interested in whether there is a feasible "orbit" or long-term trajectory with minimal burns which achieves the greatest number of visits between the two planets for a given solar orbit. It would be great if the taxi could use gravity assist at the destination planet for any course changes to the next destination. Obviously, an Earth-slingshot will not be feasible if the taxi has to get close to the atmosphere.
This "orbit" is less time-efficient than a one-shot trip because the taxi is moving continuously between the planets, and so one cannot choose an optimal alignment between them. Even so, I would hope that the average trip would not be too much slower than the alternatives. If a burn is required to shoot for the next destination, what is approximate fuel requirement?