The thing I find so frustrating about Apollo is that despite the shortcuts and compromises in design which were necessary in a crash program, the Apollo spacecraft was a very good one. I get the impression it had a lot of potential for further use and development. Skylab, as you mention, was one such case, but I understand it could have gone a lot further.
"Good" in what sense? The CSM was pretty good at what it was designed to do: ferry three astronauts and some of their cargo safely into earth orbit and in some cases to lunar orbit and back. And it was a pretty good platform for scientific equipment on the J missions.
But the CSM was very expensive, and even the part that returned to earth could only be used once. Its flight endurance was only a few weeks, long enough for a quick trip to the moon and back but its utility was seriously limited for longer missions. For example, even though Skylab had solar panels that made the CSM's fuel cells unnecessary while docked, the fuel cells could not be shut down after docking and restarted before undocking and re-entry. Each cell required a minimum load of 563 W just to keep its KOH/H
2O electrolyte from freezing. The SM couldn't carry enough hydrogen to run the fuel cells even at minimum power for an entire Skylab stay, so they shut down permanently after just a few weeks at the station. (Fuel cell power from the first Skylab CSM did come in handy before the crew could deploy the remaining OWS panel.)
The CM entry batteries could not power the entire CSM between undocking and landing, so several large silver-zinc batteries were added to the SM just to support the CSM between undocking and SM jettison.
The CSM docking system left a lot to be desired. The crews dreaded it, and it almost ruined one mission (Apollo 14). Each retraction consumed a bottle of compressed nitrogen, so there could only be so many dockings on each mission. Pyrotechnics were used for the final separation, which always seemed dangerous to me. Apparently it did to NASA too, as they required the crew to be in suits when the jettison pyros were fired in case they caused a cabin leak.
Of course much of the CSM (certainly all of its electronics, displays and controls) is now horribly obsolete, but it wouldn't be fair to hold that against the CSM in its own time. I've mentioned things that could have been done differently even at the time, but weren't because the CSM was designed quickly, for a very specific purpose, and with cost as a secondary consideration.