Because I am a living cliche, I'm writing this from a Starbucks. No wifi but I'll do a little dance to get this onto my phone and then onto the blog. It's cumbersome, but I can live with it. I'm lucky to have a Starbucks close enough to my mechanic's that I have a place to wait until my car is ready. Anyway that's neither here nor there.
I was thinking about Avatar: The Last Airbender the other day, and how it plays into my unified theory of skill levels. See, as much as we in RPGs like to have very finely graded levels of distinction between levels of ability, fiction tends to work in a slightly different way. Look at almost any movie or TV show and you'll see that competence (usually expressed in terms of kung fu or other fighting skills) breaks down into broad tiers. (I've touched on this thought before, but it's one I keep coming back to, for good reason.)
At the top is the Supreme Badass. Not every story has one of these, and if it does, they are more likely to be a plot device than anything else. This is reserved for Old Masters and Immortal Generals and the like - they simply are so good that they can't lose a fair fight (that's an important distinction BTW). If this character is a protagonist, then the story is NOT about them fighting, though it may appear to be.
Next down is the Badass. Practically speaking, he can win every fight he's in (within the reasonable bounds of genre) and fighting is his thing. Even if everyone else can fight, this guy stands out as THE warrior. It is rare that there be more than two of these in any story (one pro and one anti).
Next down are the Ass Kickers. These folks are not quite up to the badasses level but they can see it from here. There aren't a lot of these, but there are enough that they're recognizable. In term sof iconography, they are usually recognizable by their distinctive look. Even if they are members of a group, their uniform will usually be distinctive in some way.
Below that are the Warriors. They're trained , and they're better than the average mook, but they're not scary. In fiction, these guys generally represent special forces, and are often found in groups in similar (but distinctive) uniforms, such as musketeers, royal guards, templars and so on.
Below them are the Soldiers. These are rank and file fighting guys of entirely adequate skill. They tend to be faceless combatants.
Below them are Civilians, who might make a good mob or even be dangerous under a skilled leader, but aren't too dangerous.
Least of all are the Incompetents, those who are actively BAD in a fight. Cowards and comic relief tend to fall into this category.
Now, what's interesting is that this tiering resolves a lot of conflicts by itself: for the most part the guy in the next tier up is going to win the fight handily. Circumstances (injury, inspiration and so on) might temporarily shift tiers, but only so much. If someone is a tier above you, then you might be able to fight them fairly on the best day of your life, but you couldn't guarantee a victory unless it was, perhaps, also the worst day of their life.
RPG-heads can probably see the mechanic there. It may be possible to increase your tier by one step, but that would be the limit. However, it can almost certainly be decreased by multiple steps by appropriate circumstances (like injury).
Given that framework, the interesting fights are the ones that take place within a tier (and sometimes between characters within a tier of each other). Why is that important? Well, consider this: historically, RPGs are built around the idea of a fair fight (as established by D&D). Numerous rules (sweeping, mooks and so on) have been introduced over the years to try to make these fights work more like they do in fiction, but a disconnect has remained: most fights in fiction are pretty fast, with a handful of exceptions when dramatically appropriate. A model like this gives good guidelines for which fights can be simply breezed past, and which should be zoomed in on.
More broadly this model can be applied to answer one of th emost important questions of running a game: when shouldn't you roll? In most other situations, like the use of skills we have some familiarity with, we can usually eyeball the situation and determine that we don't need to go to dice. Combat and other circumstances we cannot identify with so easily tend to default to the dice, and that's not always the right choice. A simple rubric like this can really help address this.
This will absolutely not work for every game. 4E, for example, should ALWAYS go to the dice for a fight because that's why you're playing 4E in the first place: the fights are awesome. 4E also assumes everyone is roughly equal in terms of badassness, just with different roles. A game like this assumes discrepancies between characters.
Those discrepancies also point to the other things a system like this requires. If it's all on a single axis, this model kind of sucks. It requires multiple axes and different ways for characters to excel. This can be a simple as the Amber DRPG's four stats or as nuanced as any skill system you can think of.
What's frustrating is that many games fit this model descriptively, but mechanically fall short. While there aren't a lot of steps to this ladder, they are very far apart in terms of effect - much further apart than a small number of dice might suggest. 
Oy, ok, car's almost ready. This is obviously, only one edge of a larger thought, but I wanted to lay it out there so I have some context for future noodling
1 - Sometimes there is an "Elite Warriors" tier over this one if the fiction is especially fighty, or if you have multiple elite groups that need distinction, like musketeers vs. cardinal's guards.
2- Occasionally fiction respects the power of numbers to overcome a skill difference, but in more action-oriented genres, raw numbers rarely matter. Mechanically that means one of two things: in a "realistic" genre, superior numbers result in a tier bump. In more cinematic genres, the group produces a leader (or similar figure, like "The Big Guy") who fights at the next tier up. The net result is the same, but the difference in color is important.
3 - Another genre decision is whether you only do fights at the same tier, or if you do fights in adjacent tiers as well. There's a philosophical question to this as well as a mechanical one - what does fighting an underdog look like? For heroes fighting an underdog is usually a sure thing but may cost time or resources, but it might feel pointless. But when heroes _are_ the underdog, they totally want a shot at the guy. How you want to handle that suggests a lot about your game.
4 - In fact, this system is a great litmus test for any skill system. If you can't imagine these kinds of tiers within a skill, reconsider whether that skill should be on the list or if perhaps it should be rolled into another skill where you can imagine it.
5 - Fudge actually handles this decently well, but Fate less so. Aspects wreak holy havoc with this model, at least as presented in SOTC. But that's not the only way to handle aspects.