I've been reading this freely available paper entitled "A Comparison of Lunar Landing Trajectory Strategies", and have come across something interesting.
As makes sense, you can afford to lower your TLI delta-v slightly below what is required to reach escape velocity and still get to the moon: if you are still orbiting the Earth but with an apogee around the distance to the moon, you can naturally still get into lunar orbit. However, the obvious catch is that the lower you go, the longer the transfer time.
However, the lower your TLI, also the lower your velocity when you get close to the moon and hence the lower a LOI burn you need to get into orbit.
The paper claims that the total delta-v for TLI + LOI can be minimised using a flight plan which results in a 5 day trip from Earth to the moon. This is a little longer than we had been planning for (3 days), but the mass of consumables required to support an extra 2 days of habitation is very little (I would bet less than 10 kg) and the propellant mass savings that result can be considerable. The paper claims a LOI delta-v on this 5 day flight plan of 816 m/s – I have been budgeting for 1000 m/s so far in all my calculations (e.g. those that produced the graph of Isp vs total launch mass), so this is a very significant reduction which would save us hundreds of kilograms of propellant.
Of course, the huge draw back is that this is not a free return trajectory, which we have always planned to use for safety's sake. Note that this isn't necessarily a deal breaker: Some of the Apollo flights were not on free return trajectorys – Apollo 13 famously wasn't, but put itself on one using the LM engine after the SM explosion.
Obviously if we manage to get the mass of everything down low enough that we can still fly a free-return trajectory on a Falcon 9, we should do that, but if we end up going over the Falcon 9 limit and we really can't figure out how to bring ourselves back down below it, we may want to consider using one of these low delta-v transfers. It'll give us a few hundred extra kilograms to work with for a relatively minor decrease in safety.