Again, thank you for trying to see the glass as half full.
I’m pretty sure the devs said in response to some of the APG skill feats that the intent is skill feats let you do things you couldn’t do (like do something automatically instead of having to make a check) or more quickly than you could otherwise. Even if you don’t trust that to always be the case, PF2 doesn’t really use penalties to balance out feats or actions. I think allowing that check with a penalty would be fine.
They can't just
say such a thing. Where's the rule that allows it?
If you really want to hedge, then have them roll with Disadvantage. Yep, roll twice and take the lowest result. If the PC is suffering a misfortune effect, roll three dice and take the lowest result. Nothing uses that kind of mechanic (outside of misfortune effects), so you should be safe. Also, it saves you from having to come up with a reasonable modifier.
Sorry but at least in official Adventure Paths, each and every DC is "appropriate" to your level, meaning that even a small bonus totally wrecks your chances.
If you assign climb DCs on objective analysis "this cliffside is hard but not impossible" you might assign it DC 20. Adding +5 or even +10 (to climb it a special way, such as with a weapon drawn) is then fair when a mid-level hero makes the attempt.
But official APs seldom (read practically never) work that way. If the cliffside is encountered at level 5, the DC might be 20, yes. But if at level 10, it will be DC 27 (or thereabouts). If at level 15, it will be DC 34.
The point here is that any bonus or penalty can't be much more than +1 or -1 or the activity is rendered practically impossible. Meaning that saying "even without the Combat Climber feat, you can climb and fight if you make a Climb check at disadvantage" will in practice mean "you need the Combat Climber feat", since noone in their right mind will keep doing something they might only have a 25% shot at.
I don’t think you’re necessarily wrong for reading it the way you do, but I also don’t think it’s wrong to allow those kinds of things with a penalty.
Gosh, no. There's nothing wrong with this at all.
But unless we're houseruling there must be a rule to allow it. How to respond to players who feel stupid for taking feats that others don't need? And so on and so on.
I would love to be allowed to do these things, but where's the guidance?
Given that PF2 tries to leave a lot up to GM discretion
I have never seen a game that leaves
less to GM discretion. Yes, I know Paizo said they had GM discretion as a game design goal. But I can't see even the faintest trace of it anywhere. This game nails
everything down, to the littlest insignificantest thing. Just look at some of the feats - they make the smallest tweaks to the most obscure actions.
Nothing is left unregulated and thus up to GM discretion.
Also, “the system should have been more clear” is a worrisome and recurring theme with PF2.
The lesson 5E drew is that by trying to be "clear" you let yourself be drawn into a never-ending fight to provide more and more detailed clarity.
The lesson PF2 failed to draw was that 5E in a big part was successful because it rejected this approach and trusted the GM.
Everything is called a feat, but they’re not interchangeable. You’re going to be focusing on a handful of choices at any given time. If you take a multiclass dedication or certain archetypes, you might have a bit more than or up to two handfuls to consider. They could have given them different labels, but what would you propose Paizo do differently?
Not create thousands of feats?
No really. Being able to crawl or swim or climb slightly faster are things that should come - no, needs - to come automatically with a higher skill bonus. The implication of all these feats existing is that any hero that doesn't take them can't do them.
Worse, it means that monsters are strangely stunted in the most wonky of ways. (Monsters and NPCs don't take feats. Yes there is a rule saying "you can stat up monsters using PC chargen rules if you want". No, Paizo don't use it. Approximately 0% of official monsters and NPCs take levels, classes, feats and such) Example: I recently had a monster suggest athletical challenges to see who's the strongest or fastest.
My player, happy with his recent feat pick,
Cloud Jump suggested "long jumping". That idea quickly crashed and burned once we realized the monster could never jump longer than its Speed (30 ft) despite a very respectable Athletics skill bonus of +35. The DC is the number of feet you want to jump, so 20 feet would be DC 20 and so on.
You can't jump longer than your speed, full stop. The only way to break this rule is to have the feat Cloud Jump. Read the errata! It's abundantly clear that without the feat the rule says you can't jump longer than your speed, full stop.
So even if the monster wanted to aim for 40 feet (rolling a 5 on its d20) it can't aim for more than 30 feet, since that's its speed. With the Cloud Jump feat this limitation is voided, and your DC is slashed by two thirds. So the hero could go for 60 feet easily.
Making basic movement hinge on specific and non-interchangeable feats like this is exactly why I confidently say the PF2 model is irrevocably and utterly busted.
It should be trivial to realize that the solution "but give the monster the feat" is nonsensical. Nobody wants a situation where every monster must take literally hundreds of feats into consideration. This feat should never have existed. The rule limiting jumps to your Speed should never have existed. It is just one out of a thousand rules that mostly exist to justify a feat.
The game would have been simply better if all of this was just stricken from the game. Which is exactly my point. Rip out a dozen pages out of the rulebook and the game just works better. Rip out another dozen pages, and the game becomes better still.
Something is
clearly wrong with the approach with which Pathfinder 2 has been constructed.
Keep in mind that the economic reality is Paizo’s core audience wants customization.
Customization is good, but this is not the only way to go about it. In fact, of all possible ways to go about it, this is likely among the very worst.
Many are were already going to be unhappy with a system that isn’t compatible with two decades worth of content, but I expect not offering a customization-focused system at all would have been economically fatal. WotC could shift gears because D&D has history and cultural cachet that Pathfinder does not, but it also belongs to a large corporation that can keep the lights on while they’re off redesigning the system.
I disagree. "Shifting gears" is just about the best praise 5E could be summarized as. 5E managed to reinvent the game, shed a truckload of old rules deitrus, and move on to huge huge success.
Despite Paizo having access to 5E for five years, they were unable to replicate any of it. In fact 2E doubles down on exactly the polar opposite of what 5E so very clearly have proved beyond a doubt what gamers want.
Honestly, I hate skill actions. I didn’t like them in the playtest, but I tried to give them an honest shot in the final game, but I ended up disliking them in the end for the same reasons I disliked them in the playtest. It’s not that they constrain creativity. They’re problematic because you normally only use a few, and then when you use one of the other ones, you’re like: “Okay, hold on. Gotta look this thing up.” If you’ve got Archive of Nethys on a laptop or tablet, you can be quick and hide it while people are rolling dice. You shouldn’t need an app or a website to run a game. It is incredibly irritating that even after 20+ sessions I still haven’t fully internalized them.
The skill actions themselves are slightly fiddly, yes, but they're not even close to where the real problem lies.
Sure I don't need a separate Balance action from a Tumble Through action from a Squeeze action, but that's a minor annoyance compared to what I'm talking about: that you might be Legendary in Acrobatics but you would still squeeze
5 feet per minute 
unless you have taken
exactly one specific feat out of the several thousand that's on offer.
I’ll totally concede that this aspect is really fiddly. They enumerated too many things. They got rid of the myriad of modifiers from PF1, but they should have also done the same to the various uses (instead of codifying them in freaking actions!). I wanted to believe that they were a part of the modularity, but they’re more like modularity gone wrong. And I was wrong about them.
I would say that for each PF1 modifier or exception they cleaned up, they added two new modifiers or exceptions. There is nothing simple or easy about PF2. It's insanely cludgy and fiddly and hard to remember and hard to calculate and you roll endless d20s and other dice...
This is what I would do to fix them, and I may incorporate some of this in my game. First, I would move the combat maneuvers into the combat section. Put everything together, so you can actually find it using the book. Second, get rid of almost all other skill actions and activities. Replace them with a simplified skill check. In essence, if you are in a situation where both success and failure have interesting results, roll to see how it goes.
This probably invalidates a swath of skill feats. Is that a bad thing? It would make it abundantly clear that skill feats were not meant to stomp on skill checks. You’d probably need to look at a handful of situations where the actions are still wanted or needed. Treat Wounds probably needs to go in the exploration chapter as an exploration activity.
But Paizo can't do this? How will they keep selling books if they can't shovel new feats out by the hundreds...? /s
