Yeah, i just looked after that in the book and yes: It seems like you are right... you found ANOTHER unreasonable decision he made while writing that book. Even if they would have needed more fuel in the CSM to support a landing, then get to the station before finally leaving for earth: It would have been doable. There are three options jumping into my head to solve such a problem: 1st: Don´t use the CSM for the landing at all, instead they should use Moonlab as their staging point, yeah: That could mean that they need more fuel for the lander, but i think that it should be doable (especially the computers they can build into that thing would be much lighter then the ones in the original LM-production-run. 2nd Option: Modify the CSM with larger tanks (i repeat. Lighter computers and probably a lot of the other stuff too, and there is no need for experiment packages on the CSM, so there is possibly more space for fuel tanks. The science-stuff can be done by Moonlab, that´s officially the only reason for it´s existence (yeah i know: Shooting that thing to lunar orbit was more of a PR-stunt then anything else), or finally option 3: Use an unmodified CSM, support the LM-Landing as it would have been done in a normal Apollo-mission and then fly on to the space station. When the Cargo-Module/ the LM-Ascent-Stage is docked there, the CSM is refueled at that station.
(sorry for the mess)
having a lunar space station is STUPID, there's no magnetic field, so you have to bring that with you, its hard to get to, limiting mission accessibility, and most importantly, whats your backup if the csm engine fails, the next mission is a year away)
honestly, I think he left it out for the "the book is long enough, and it isn't about the moon", but with lighter computers, additional fuel can be held (by modifications), and even assuming a direct to-Moonlab trajectory would leave enough fuel for a landing (CSM being the active portion of the rendevous, the LM just being transferred).
Now, we need to know the ORBIT OF MOONLAB, is it Polar or equatorial
a Polar orbit would put way more stress on LM capabilities, especially with the orbital Plane problem (Orbiting CSMs on apollo missions had to do this regularly to keep "flying over the landing site", a polar orbit would make any such landing impossible without a "super lm"
Equatorial orbits would be EXTRAORDINARILY limited in their landing sites, likely Marius hills or other low inclination orbits (again depends on how equatorial we are talking 10 degrees or 0 degrees)
each of your approaches is good in its own way
The moonlab-based approach is best for an Equatorial landing (little plane changes and saves a heck ton of fuel.
Larger tanked csm is best in a polar orbit, having the LM land and the CSM do plane changes to keep the orbit correct (and trajectory planners could "fix" the orbit to be in Moonlabs plane at the end)
this would work even better on equatorial missions, having the margin of fuel being much more than a regular csm, and assuming NASA has the budget for modifying its remaining LMs (14 and 15s H-Mission, and 16 and 17s LMs (the book referencing 16 confirms they were at least "built")) they could attempt a 4 day mission
The Martian has this plane problem with its abort, the site is 49 degrees, and then the "shoot the dude into space idea", which is a complaint people have (the first is, they abort at day and launch at night, presumably when the launch window is up, and the second would have had years of planning to do)
Watney even describes a training program that had them in the same room for a few weeks, to simulate this (Hermes not having the shear thrust power to grab them immediately, taking time to match planes with their MAV
a regular CSM/LM mission would be usable on both, more suited for Equatorial landings (less tight fuel margins), polar orbits would be extremely finicky and be tight on fuel (trajectory corrections, orbit insertion, descent orbit insertion (if its not cut), plane changes, and moonlab rendevous and docking, and most importantly, Trans-Earth Insertion)
an uprated CSM would do wonders in this regard
so here I go (plan across all three options being friendly to the book)
the first two moonlab flights would build up experience with lunar operations
the third would bring an LM to land, a J-mission LM, (polar orbit would be lightened, having a 2-day stay time or no rover)
the fourth would bring an uprated H-mission LM, which would land
the fifth and last moonlab mission would bring the last J-mission LM to the moon (polar orbit being lighter, less stay or no rover), after which the Moonlab program would be shut down due to Apollo-N and the reprioritization on Mars missions
that gives two "learning flights" and three missions and is the most likely scenario i see
I would say have the second moonlab flight be a landing, but after the outfiting on Moonlab 1, the prority would be getting it up and running before any attempt at a lunar return, hence the third mission having a lander (maybe Moonlab 2 uses a Lunar Module as a cargo carrier)
the extremely stupid idea is having said cargo lm be a fully capable lm, being dormant on the station, when moonlab 3 arrives, it immediately lands, does its stay, and returns to moonlab, where after a month, a second landing occurs with this second LM, being more bang for your buck type thing
in terms of logistics
a dedicated cargo carrier/fuel carrier to the Moon is
stupid like wow that's really stupid, stupid (not being mean sorry, like really sorry. sorry), the shear mass requirements for that with what they had at that time would be expensive, it would require 30 tons to LLO or PLO, which would be higher than most rockets capable of the time, and have to be used on a Saturn V booster (which is why Baxter had them use the old LMs as logistic carriers, it kinda makes sense in that regard [two birds with one rocket]), yes it would work at allowing access to the whole surface (nasa limited sites on safety resons, rightfully so), the problem is the interesting sites that scientists would want are still hard to get to, the poles are just asking for trouble, Tycho is a deathtrap (the boulders), and the far side is still, the far side, so your looking at a new comsat constellation in lunar orbit (the book never said they had communication when going aroud the far side), which would add furthur cost\
if moonlab is in a polar orbit then this "fuel carrier" is uneeded, it would only be needed in a non-polar orbit
the Moonlab-based approach with a cargo carrier would work best here, having three birds with one rocket (station cargo in the LM, CSM, and Lander)
if I was in charge of planning, I would have canned 17 and had 16 land at Descartes with an LM stocked waiting for them, with a stay of 2 weeks, would put all missions to shame, dunno if Duke would be canned, but maybe Schmitt got moved to 15
with Skylab-A and -B, we don't get much info on them at all, to say nothing of logistics or operations (we know some stuff not all), I would assume they have a variation of the Aardvark or something, to be used in regular operations, either launched by 1b or Titans
In Voyage, probes still happen, only being Mars focused, with Voyager being canned in favor of more Mars orbiters and landers, so the tech would still be there, just not as well practiced to the same high-stress extent as otl (some of the voyager problems were nuts)
and his canning of probes did happen in real life, due to the shuttle, NASA canned most if not all probes after 77 to the mid-80s, where Challenger then pushed them into the late 80s to 90s, leaving a nearly 12-year "gap" in probes launched by the US
Startracking would probably be on the same level as otl, probably being a duel system with a computer and human o[erator