[jhgb]

There's no reason why you couldn't use an accelerator for transporting people. The velocities are fairly low so acceleration can be moderate as well. Also there's no air to complicate things with aerodynamics, so all you have to do is to design a high-speed maglev with auxiliary thrusters and RCS that could lift from the tracks at orbital speed and then land back on them again (which is much easier in the complete absence of atmosphere which turns it into an orbital mechanics problem).

[inglor_cz]

So I put the numbers into a calculator. Let us say that 3G for a minute would be tolerable for most humans, at least those that venture onto the Moon.

You need approximately 60 seconds of 3G acceleration to reach orbit around the Moon. That would mean some 54 km of an accelerator. That is more feasible than I instinctively expected. The main question is how much metal would be needed for 1 m of such accelerator track.

[jhgb]

The real question is how much material you'd need for a ~80000-km-sized vertical structure that the lunar elevator is expected to be. The ~1:1000 factor in the largest dimension required can't be good for a practical elevator.

As for the track, I suspect it might be (order-of-magnitude-wise) similar to existing maglevs on Earth. On one hand, you need to reach higher velocities, higher precisions, etc. On the other hand, any supporting structure can be substantially more lightweight (much smaller gravity, no corrosion resistance required, etc.).