A solar wind-derived water reservoir on the Moon hosted by impact glass beads
nature.com
nature.com
That's the same as a sphere of water with a radius ranging from 415 metres to 4.01 kilometres.
https://www.wolframalpha.com/input?i=volume+of+fresh+water+o...
Given the history of past impacts and maybe even other solar system events contained in that soil perhaps it is not worth it.
It's nice that it recharges on a 15 year timescale though, which makes it something of a renewable resource.
Water is H20 by composition, but since Oxygen atoms are at least 8 times heavier than Hydrogen (and usually about 16 times due to neutrons), the Hydrogen is only 11% of the water by weight.
Hence per Kg it's far more efficient to split the water into hydrogen, and re-oxidise it with native lunar Oxy; 1kg shipped water could be 1kg Hydrogen that would produce 9Kgs water once oxidised (ignoring difficulties in Hydrogen containment).
I'm pretty sure that we can do better by just using the energy used to split the water to accelerate it instead.
Here's an example of Nasa using electrolysis + water for maneuvering cubesats: https://www.nasa.gov/feature/ames/ptd-1
I don't know why I didn't consider h2o molecules as ion engine propellant.