Pathfinder 2 is absolutely everything but "consistent with few exceptions". The game in general is shock full of weird little special rules, conditions, and limitations. Very little of it feels earned, or contributive to the play experience.
The game screams for a general mechanism to grease the wheels, one that empowers the gamesmaster to allow slight deviations from what the framework lets you do.
I actually recognized that in one of my other posts when I brought up skill feats. I think it’s a good idea, but it screws up the simplicity.
When I say the framework is consistent with few exceptions, I mean the
framework. You get three actions and a reaction, and you don’t have various exceptions written into the framework. We don’t need to have discussions about how many times a paladin can smite in a round or whether under some circumstances we can cast two spells because the action economy handles that. We also don’t need to memorize a table of situations where something provokes an AoO, but we also don’t have to give up the richness that 3e and PF1 had by simplifying it down to one or two events. Yes, there are traits like [Flourish] and other traits, but if you’re not dealing with those, you can ignore them. If you are, they’re explicit about what they do.
The same goes for making rolls. Everything is a check. Everything works on a same scale. If you can justify rolling a Reflex attack versus an Attack DC, the math will work. Something modify Strength checks? Then yes it affects your attack roll. They restate that in conditions like enfeeble, but I think that’s just to accommodate people who are used to attacks and saves and checks all being distinct things.
I use this in my exploration procedure when trying to force march: You make a Fortitude save versus your Constitution DC. It’s succinct, and it lets me avoid having to try to write something that understandably conveys the same thing but in more words. “Make a Fortitude save with a DC equal to 10 + your level”, but I’m pretty sure people will sometimes forget to add their level. Roll a number on your sheet versus another number on your sheet is much easier, and it feels more flavorful.
So when a situation occurs, I can use the framework the system provides to adjudicate it. If someone is doing something with their hands, and it’s not just an Interact action, then I can determine that it has the [Manipulate] trait, and things that key off that all just react accordingly.
For example, I have my PCs carry bows in slings when they’re traveling. How long does it take to ready a bow? You need to Interact to take it off your back, Interact to remove it from the sling, Interact to brace it behind your leg, Interact to bend it forward, and Interact to put the string in place. Five actions! The system doesn’t have rules for that, but I was able to come up with something that just naturally fit in the action economy
and actually makes sense realistically (just go watch some videos on Youtube and see how long it takes people to string a bow). No rulings, just applying the framework.
Like I said before, skill feats are problematic. They muddle the framework. I don’t find your example problems compelling, but I agree with the basic premise: it should be possible to do something unusual with a skill at a higher DC than normal. If you want to Make an Impression to a group, it should just be possible to attempt that at a higher DC. Technically, that wouldn’t negate the benefit of Group Impression, but the rules aren’t clear on being able to do that, and we can’t trust that Paizo will never design skill feats that don’t mess up that approach.
Instead of looking at their prototype three-action framework and going "we need to let the GM allow skill checks to transcend and ignore the weird artefacts that can happen" Paizo instead said "let's double down on the hard no's by inventing a feat for everything we can identify as something to lock down!" "Gating things behind feats must be good, right, since it means more options! Right? Right?"
In theory, yes. Having non-combat customization as its own thing ensure those options actually get taken. Players have a tendency to focus only on combat benefits. In practice, see above.
Everyday actions like climb, jump, crawl are weirdly locked down. Related feats come across not as making you awesome. Instead they make you not actively suck.
The GMG explains why: they wanted to make it easy to understand how things work. It should have been said in the CRB, but they
recommend allowing PCs to combine types of movement into two-action activities as appropriate. Of course, we need to trust that any skill feat written in a future that confers a similar benefit must be written to only take one action, so it doesn’t negate the thing we thought (and they told us!) we could do.
Nearly every feat and item brings some kind of tiny puny and frankly unnecessary special condition or limitation that increases the rules burden on the player and GM.
Far too many feats work the "same, same but different" way than a feat you'd think would do the same thing. It does, just with niggly little differences.
Too many magic items work like in 4th edition in that they're too much of a hassle to be worth the bother. Getting a +1 or +2 bonus once every blue moon on a specific check just isn't worth remembering.
I could probably present a hundred different examples proving this beyond the slightest doubt, but I fear it would mean going down a rabbit hole that gives actual mental damage, so I shan't.
I’m just lumping these all together because they seem to go to the same point: there’s a lot of stuff, and it’s not interesting enough to justify its existence.
I think it was a mistake not to have a mechanism for combat styles. All the martial classes should be able to pick a style and have feats available to take from it. I’d feel comfortable letting martial classes take a fighting style dedication provided they spent their 2nd level feat on a 1st level class feat, but that’s a house rule. Anyway, if martial classes could do that, then instead of having things like Twin Takedown and Double Slice, which seem superficially similar, you could just have a ranger thing that made your Double Slice take only one action when you are fighting your hunted prey.
Little bonuses don’t feel good initially, but players eventually start to internalize that a +1 means more than just an increased chance of success. Is there a way we can dig into your complaint without having to go through lots of examples (since you don’t want to go down that rabbit hole)? Is the issue that they don’t allow for enough character distinction? That if I take a feat to let me specialize in climbing walls or fighting ninjas, I’m only “a little” better instead of substantially better?
We’re probably not going to find common ground on consumables, so I’m not sure how far it’s worth going into it. I think there are situations where consumables are useful. If continuing to rest has an opportunity cost, then elixirs and potions are useful as a way to avoid another ten minutes of downtime. If encounters are foreshadowed, you can prepare for them with the right talismans. If a different game where those things aren’t true or don’t happen, they might be less useful. I’d posit that perhaps that style of game isn’t an intended way of running the system, but it’s also easy to just ignore the things that don’t work and replace those things in treasure with items that are actually useful.