You might want to take a look at this
thread ... turbo lift cab depots Franz Joseph style (other than those for repair and maintenance) would be a waste of space
Okay, I read the entire thread (thanks for the link). Plenty of interesting theories there, but the idea that dedicated cabs "follow" key crew members is problematic. It might work in TNG times, but in TOS the ship does not "know" where individuals are. You can scan for life signs and try to pick out a Vulcan among Romulans, or eliminate heartbeats until you can deduce that Finney isn't dead, for example, but the ship does not "know" the wherabouts of the captain until he punches an intercom panel or something. (And even then, it still might not. I don't recall any evidence of the ship's computers maintaining this kind of information, and in fact that plural — computer
s— reminds me that in the 1960s TOS, there really isn't even a concept of a "central brain" running the ship; that concept would have to wait for TNG and audience familiarity with networked computing.)
Anyway, maybe I'm missing something here, but I don't see how avoiding cab depots saves any space, really. If you don't have "spurs" or garages or some other dedicated space that's not needed for actual cab routing, then the alternative would seem to be the idea that all shafts are double- or triple-wide to allow cabs to "pass" each other — and that seems
massively more space-wasteful to me than a few depots sprinkled here and there.
The fact that we never see anybody waiting for a cab — turbolifts open instantly 99.9% of the time — suggests that the number of cabs at least equals the number of termini (or the number of
routinely-used termini; you probably don't need predictive, instantaneous cab availability when leaving the Christmas ornament storage lockers). Now, possibly the ship could continuously monitor life-signs and only maintain ready cabs at termini in sections that are actually occupied; but based on the typical hustle-and-bustle activity depicted aboard the
Enterprise, it seems safe to conclude that the majority of sections are usually occupied anyway.
In addition to the baseline number of cabs presumably being somewhere around the number of termini, we need extras to account for peak usage (shift changes, battle-stations calls, evacuations...), maintenance downtime, decommissioning, battle damage, etc. Surely the total number must then
exceed the number of termini; and all those cabs need to park somewhere. If where they park is defined as opportunistically "anywhere" in the shaft network that makes sense at the current moment, then this means the entire network needs to be constructed with major redundancy to accommodate passing, rerouting,
and parking. I recognize that this still could very well be the best answer; but it sounds far less space-efficient than a small number of dedicated garages where cabs can be packed together out of the flow of traffic. And is it even feasible to implement your deck plans with double-wide shafts everywhere?
Every time a cab leaves a terminus, another cab needs to replace it (and promptly, as in the Lazarus-followed-by-redshirt example). Where does this replacement cab come from? We can't just steal it from the next-closest terminus; that only creates a cascading vacancy effect that is wasteful of energy, increases the number of cabs in motion at any time, and complicates overall network logistics. Far simpler to deploy a replacement from the depot nearest the departure point, while sending the evacuated cab from the destination point to its nearest depot.
Of course, whenever possible, for efficiency's sake the routing algorithm will try to take advantage of localized "ionic" relationships, or complementary needs, where one terminus must get rid of an empty cab and another needs a replacement, and a reasonably short path exists between them that will not impact other occupied cabs in transit. But that's an ideal case, not the norm: the baseline algorithm will need to be able to count on always having a place to dump a cab that needs to be gotten rid of, and another place from which to fill an empty terminus immediately. If we don't have garages or double-wide tubes, then what's the alternative?
One way or the other, the network needs more space than just the baseline pathways, don't you think?