Tracking stuff like that varies between groups and games. Heck, how much I track varies depending on the theme of the campaign.
Here's an RPG principle: zoom in on whatever is a significant, immediate conflict for the main characters, and handle it in fine detail. Everything else, handle with less granularity of resolution, or just handwave.
1E D&D assumes that fights with monsters or with rival humanoids will be a significant, immediate conflict for the main characters, and so it has lots of rules for what EXACTLY happens when you attack someone with a bec-de-voulge-guisarme and they're wearing banded mail. (There's a table with modifiers for every weapon, vs. each type of armor, in the original DMG, so that plate mail is great against swords but less so against warhammers. I never saw anyone use that table.)
It also has NO generalized core rules for perception checks. There's a rule specific to finding hidden doors, with a modifier if you're an elf. There's a skill for finding traps, which does not interact at all with your chance of spotting a hidden door. And in the monster manual, under Yeti, there's a rule specific to spotting yetis hiding in snow. Because hiding on a same-color background is done only by yetis, and not by any other species, natural or otherwise.
Starvation, which I'd bet had killed more humans than armed conflict, at least up to Elizabethan times? Not a significant, immediate conflict under consideration, which makes sense in context. So no rules.
If *all you care about is resolving the outcome of fights*, this is appropriate game design. If you want three pillars, then you shift the rule design accordingly.
A want a set of rules which supports BOTH handwaving AND close-in, fine-detail resolution, on any topic, depending on the themes and specifics of the story at the moment.
When the story is about finding the Amulet of Plot Significance, which is buried on a mummified corpse under a pyramid, then you take a caravan from Starting City to Pyramid Location, it takes you a week, got any specific plans for that week of travel? you're practicing the flute and chatting up the dwarf? fine, so noted. Foreshadowing is established for the pyramid trap with a musical solution, and for the moment when the dwarf figures out the sloping tunnel trap and can only save one of the PCs. This occupies one week of the character's lifetime - and half an hour of player's session, at most. The first hour that the characters spend in the pyramid, however, is played out during two hours of the player's session time.
When the story is about solving a mystery when one of the NPC caravan travelers dies (Murder on the Silk Road), okay, THEN you play out that caravan journey in detail, day by day. Camels, dromedaries, horses, mules? Are you riding them, or leading them while on foot? Are you following a river, or going from oasis to oasis, hoping the water doesn't run out? Who's watching out for bandits, and are you leaving a trackable set of footprints and hoofprints? Are you wearing heavy armor all day long, and what's the exhaustion cost for that?
Spending an hour of session time on those topics, in the Pyramid Amulet Quest, is not worthwhile. But when days 1-3 of the caravan journey are setup for what happens on day 4, then those topics ARE worth an hour, because they affect who's where, and who has whose trust, and who's ready for combat, at noon on day 4.
Same rules system, same gaming group, same campaign, *different story*.
Which means that rules for the specifics of caravans should be *available* and not *mandatory*.
Bonus points if you can apply this principle to spending gold on magic and on nonmagic items!
