As a way to get stuff off the Moon, how about a hammer-throw type surface based spinning tether lunar launcher (suggested by "Robert" at Selenian Boondocks about a year ago), a 200km radius of rotation gets 50 tonne manned ship to escape velocity
with a centripetal acceleration of 28m/s (2.8g), rotation speed of 6.8 rpH, at 50GPa strength the tether weighs only about 15 tonnes.
To decrease the time between launches, you could have the
payloads travel down a permanently extended tether.
When you get to this radius of rotation, the possibility exists for the tether to catch
payloads from space, making low g rocket free transport available in
Using the mass driver you face a major problem in power storage and
rate of discharge, the solution is usually to launch small
payloads, but using the "hammer-throw type surface based spinning tether
lunar launcher" the power storage issue disappears, the tethers rate of
rotation can be increased as slowly as power supply allows, and the high
payload/ tether mass ratio allows large individual payloads.
A tether with a radius of 4km located at Earth-Moon L2 would catch payloads arriving at
270m/s, subjecting them to a centripetal acceleration of 18ms^2, tether
rotation is at 3.86 degrees/sec. A carbon nanotube tether for a 50 tonne
payload/manned ship need only weigh about 200kg(!) at 50GPa strength,
so would need to rotate around something substantial, EML2, as you probably know,
is unstable, so the momentum of the arriving ship could probably be
used to maintain the receiver at EML2 if positioning at capture was
done with that in mind.
Energy costs for the tether option at 10c/KWhr are about $75/tonne assuming negligable losses, perhaps $100/tonne including losses.
Update 8th June: Well this idea has been looked at by the experts after all, it's covered towards the end of this document.