My experience differs sharply. A short rest is pretty easily available any time that the party doesn't have an imminent deadline. The total time cost is low, or even almost zero if the party just stops for their next meal earlier than originally intended. Moving lunch up a couple hours, for example, doesn't change how much the party can get done in a day.
By contrast, spending 9-23 hours for a long rest (depending on the remaining time before the party is eligible to take another long rest) slows the party strategically. If everything they do takes significantly longer thanks to stopping early for long rests, they won't accomplish nearly as much.
I suspect the difference in our experiences may come down to how many active quests/opportunities/goals/priorities the PCs are pursuing simultaneously. In my campaigns it tends to be half a dozen or more at any given time, which creates background time pressure. Even if none of those priorities have an explicit doomclock, time spent on one priority is time not spent on the others, and (almost) none of them are so static that they'll wait around unchanged for the PCs to eventually get around to them.