So the funny thing is I have tried the same thing in reverse (turning the short rest classes into long rest classes). I figured it would be easier because I only really had to worry about 3 classes instead of 9. The results of that experiment were me learning that It doesn't really work that great (Increases the tendency of my Party to Nova the crap out of enemies. Also I had two monks so any boss got stun locked to death.) While I doubt turning long rest abilties into short rest ones will have the same issues I do suspect it will unbalance the game and make it less fun.
I think a simpler solution to the problem is to mess with the rest mechanics rather then trying to change a bunch of classes. I've been trying 13th age and am one hundred percent stealing the idea of "you get a long rest after about 6-8 encounters not whenever you sleep." In addition give each person only 2 long rests per short rests and it's done.
So my plan is to turn this upside down.
First, go with gritty rest schedule (with some changes). That lets me control the frequency of rests
narratively.
An "encounter" is a bunch of stuff you have to mop up in seconds, minutes or hours. (an encounter is "multi-part" if there is more than a minute gap between pieces)
A "scene" is a bunch of stuff you have to wrap up on the order of hours. If you have time to take a nap and come back with the problem unchanged, it is
not a scene.
A "chapter" is a bunch of stuff you have to wrap up on the order of days. If you have time to go off and take a week break in civilization, it is
not a chapter.
Now build adventures from these components. To keep things simple, I'll build them based on
tiers, and a given chapter/scene/encounter can be hard or easy within at tier. So T1, T1.5, T2, T2.5, T3, T3.5, T4, T4.5 and T5 (just because). And maybe T0.5 as well.
An "adventure" is structured as a bunch of chapters.
We now lay out a chapter. Chapters are built out of scenes, which I can describe roughly and give a tier-level to. How many and how tough those scenes are determine how tough the chapter is.
A scene could be "kobolds in an ancient ruin", "highlanders guarding a mountain pass", "pirates raid the docks", "griffon whose mate is sick", "undead monstrosity wandering the wilderness", "secret tunnels under the inn".
Those elements need not be linearly related to be in the same chapter. They can have connections to each other. So long as the entire chapter has
something happening when the PCs say "well, I go off and read books in the library for a week", and some connection, they are connected enough. "Leftover scenes" can even be consumed in other chapters (or recycled, hehe).
Next build the scene. Scenes are sets of encounters. The difficulty of the encounters, and how many of them, determine what "tier" of scene it is. If I'm building the scene difficulty from the top-down, this gives me a budget; if I'm building it from the bottom-up, the encounters determine the tier of the scene. Not everything you
can do in a scene need be part of its budget; an ancient red dragon offering a deal to PCs is not required to be fought, so using it as part of the scene's tier is misleading.
Much like chapters, there should be a consequence if you walk away from a scene and go and and have a nap. If you raid the kobold temple and you leave half way through, the kobolds are not going to wait around. Maybe they'll rush the ritual they are doing (with interesting consequences), or maybe they'll flee and come back in a few weeks with reinforcements, or maybe they'll just kill their hostages and flee.
The consequences of engaging and then walking away is what makes a scene a scene, and not just a bag of encounters.
Scenes that are just 1 encounter are perfectly ok; just don't over use them.
Then the encounter.
The encounter uses any of the many encounter-building mechanisms. I find it useful to map it to either an encounter level (a party of 4 PCs of level X would find this to be a medium encounter) or a tier level (which is just a simple remapping). I use a bit different math than the DMG system, one where I can add up a number for each PC (basically the PC level, with a +1 bump at 5 and increasing by 2/level at level 11 and over) and for each monster (similar, but uses CR instead of level, and increases even faster after 20) and simply compare the ratio of the sums. Much less fiddly than the DMG system (or even the UA one) and gives very similar results.
---
Playing with it, lets say we have a chapter about (say) a red dragon demanding tribute. And the locals are willing to give half of their 1/10 years tribute to you... if you slay the dragon. (they ran short, and know what happens if they fail to pay up).
That is a chapter.
Scenes could include
(a) In-town, gathering supplies information and transport. (wasting time here means the dragon could arrive before you kill her)
(b) Something that happens while getting to the Foot of the Mountain (dangerous wilderness) (trip takes 2-3 days, scene happens on one of them) (optional; some travel modes could skip it)
(c) The Kobold Temple of Worshippers (has a tunnel into the dragon's fast) (one possible way in, either discovered in (a) or found by "kill the kobolds" strategy)
(d) The Back Way up the mountain (environmental, maybe dragon ambush if the players take too long) (another way in, either found in (a) or via exploration)
It is unlikely that PCs will do both (c) and (d).
(e) The Dragon's lair (the dragon herself, and some guards are possible scenes) (engage the lair and retreat, and the dragon goes and burns the town down)
Failure on (c) or (d) could result in the Dragon showing up and fighting them outside, or even going after the town.
Not all of these are "must do". But the Dragon is coming for the tribute in 3+1d6 days, so taking a week break is probably inadvisable.
Each scene can be avoided or skipped. As this is obviously a T2 or T3 adventure, either the mountain is warded against teleportation effects (if it wasn't, the non-Dragon-fight parts wouldn't be scenes in the chapter, as they are irrelevant to the chapter; you'd have to have a different chapter goal and initial state for them to stay in).
I can then go and plan out each scene.
---
Under this model, the players are
encouraged to take a break and rest up whenever they are hurt and need recovery. This is only reasonable. They should only not do that if they have
reason to; the scene and chapter structure is designed around giving the players reason not to go off and rest, and requiring the DM (me) to make that plan explicitly (and why the players would know about it reasonably) is part of my job.
The day-week rest structure makes building such narrative reasons far easier.
I can even make a free-form act; in 2 months, the regency ends, and the prince ascends the throne. Then I can toss in a whole pile of scenes about factions trying to kill the prince, make the prince go mad, kidnap the prince, whatever.
Some of the scenes (static ones) can be ploted out in space with minor time evolution. Others can be plotted in time and space (on day 10, the assassination of the Vizier occurs). Or I can toss out a bunch of factions and have dice and players actions determine how they advance, something like dungeon world's dooms.
For toy mechanics I haven't tested, each faction has a power level measured in a die size (or count) -- like d10. Every week a faction rolls their power, and adds it to where they are in their plot.
Factions have a
plot. This plot has milestones. These milestones are things that happen that PCs can notice, and can have effects (like, if unstopped the faction gains power). Factions can even be killed (this can cause other factions to gain power; the Vizier might be hindering the Bishop and the Regent, and if the Vizier is eliminated the Bishop's power die grows by 1 step and the Regent gets a 2nd die).
Plots should involve setup (something the players could detect) and execution (an event that, unless the PCs are right there, might not be able to do anything about). At 30 points, the Bishops' assassins attack the Vizier; roll Bishop power +3 against Vizier power to see if they kill the Vizier. On failure, the Bishop power die shrinks 1, on success "Vizier is eliminated" occurs.
Before 30 points, things could happen to let the PCs know that something is up; but if the PCs didn't stop the Bishop, the murder attempt happens. If the PCs happen to be with the Vizier, or close enough to intercept the assassins, their actions have the natural consequences.
The PCs become able to take breaks whenever they want, but as they do so the plots advance. They can get recruited by various parties, or act on their own against the plotters. The consequences of inaction are planned; their actions gets to break those consequences before they happen.
