I don't like playing in NWP games. Players seem to fixate on finding ways to use their NWP that aren't very organic.
Yes, that can happen in any game with a skill system. However, I have found it to be the easiest bad habit for players to break, given the right incentive. I explain to my players that for many actions, the skill system is a safety net. If they can figure something out methodically by question-and-answer with the DM, they may succeed without having to take the risk. For example, if the players puzzle out how a trap mechanism works and figure out how it can be disabled, they don’t have to make a Thievery check.
I do this because of an article I read, I don’t remember where, which said this was the intention with respect to the Thief’s abilities. The intention was that by default traps should be discovered and disabled by player skill, and the Thief’s abilities were really intended as a sort of Hail Mary pass. Which is why Thieves are so crappy at some many things at low lever; you aren’t supposed to be relying on them in the first place.
Note I don’t just use this for traps, and finds that it works for many other skills as well. And players looking for an edge (aren’t they all?) generally get on board pretty quickly. It helps if the stakes are high and the DCs are difficult.
As for rangers or any other class, I let them do anything that is in the archetype for the class, using the roll-under rule on DMG pg 110:
There will be times in which the rules do not cover a specific action that a
player will attempt. In such situations, instead of being forced to make a
decision, take the option to allow the dice to control the situation. This can
be done by assigning reasonable probability to an event and then letting
the player dice to see if he or she can make that percentage.
So if you're a ranger and want to climb a tree, your percentage is probably 100% if you have a rope for primitive fall protection, the tree is healthy and of a type good for climbing, and you have no time pressure. It might go down to 50% if someone is shooting arrows at you, the tree is dead with compromised wood, and there's only a couple of rounds before the branch of someone you're trying to reach, breaks.
I see a couple of problems with this. The first is the statement in the quote that this applies where “which the rules do not cover a specific action”. The problem is, the rules DO apply to the actions we are discussing. They grant the Thief an ability to climb walls, and they grant the Barbarian an ability to climb cliffs and trees, but they do not grant that ability to the Ranger. Every group I ever played with interpreted this as meaning only a Thief could do these things, and from my interactions online I don’t think this is uncommon.
Another problem is that there is a risk with
ad hoc decisions of this nature that you could accidentally grant a better chance to the non-Thief than the Thief gets.
Say a ranger wants to climb a cliff in a natural setting. The cliff is dry and rough, with many projections; in other words, it is pretty much like a modern climbing wall. A DM could easily make an
ad hoc decision that a ranger ought to be able to climb that without bothering to roll. Maybe you wouldn’t make that call, but it is easy to see that someone might.
But per DMG p.19, the Thief has the same chance to climb a dry surface that is “rough and with ledges or many projections” as he has to climb a dry surface that is “very smooth – few cracks”, the only difference being the rate of speed. For a 1st level Thief, that is 85%, less in leather armor.
And your point only speaks to climbing. Stalking skills are still absent from the ranger, whom I imagine to be an accomplished hunter. Surprise is a limited proxy, since there are many instances when moving silently or hiding are useful other than when you stumble into an encounter.
Moreover, you don’t seem to do anything all that different from what happens with a properly adjudicated skill system. When a player declares an action that is something you think the PC ought to be able to do, you decide whether it is something that is easy enough to be accomplished automatically, or if there is risk. If there is risk, you decide the matter by the roll of a die, either by using a roll-under mechanic or by making an
ad hoc decision as to the probability of success.
With a skill system, if the player declares an action that is something that the DM thinks the PC ought to be able to so, the DM decides whether it is something that is easy enough to be accomplished automatically, or if there is risk. If there is risk, the DM decides the matter by the roll of a die, using the established skill system and determining the probability of success using established guidelines.
The only difference is whether, when there is a risk of failure, the probability of success is determined systematically or in an
ad hoc fashion.
1E has secondary skills, which aren't directly applicable to your question but do show what the design wanted to granulate and what it didn't.
4e digression, which is probably relevant to 5e: I do something similar in my 4e game using backgrounds. I assume the PC has proficiency in matters related to the background from which they have chosen a background benefit, and possibly others if they seemed to have developed them somewhat; it is a bit
ad hoc.
Conversely, I assume that they are not proficient in matters unrelated to their background or to adventuring in general. The assumption is that adventurers are trained at adventuring, so that, for instance, a ranger can’t use the nature skill as a proxy for farming unless he has a farming background. This is actually a diversion from the rules as written, but I prefer it.