Jon Goff has a proposal for doing lunar missions with a bi-elliptic transfer. It makes a lot of sense, actually, for Low Lunar Orbit, or deep space, though I’m not sure there’s any benefit for a Lagrange point, because the plane change out at that distance doesn’t cost much anyway (one of the many reasons that I find Lagrange points preferable to LLO).
The basic idea is to do the plane change at a very high altitude. In fact, this is a technique that can make sense even for LEO plane changes, if they’re big enough. I forget where the crossover is, but there is a certain amount of plane change where it is actually cheaper to go out to GEO (or higher) and back than to do it with a single burn in LEO. We looked at it a lot back in the eighties when we were doing tug studies.