Are GM's (or players) interpreting die results just by the final scores, or are they looking at which of those final scores are on what type of die and applying more detail?

What the frizz do I mean?

You are attempting to fix your landspeeder during a sandstorm and the GM applies a setback die. You get a success on a green and a failure on a setback, failing the task. Has it been your practice to simply state you failed to install the hypercore or do you say, since the failure was on the setback (applied due to the storm), sand gets into the hypercore, fouling the engine, providing detail based upon what die provided what?

OR

You try to bash the Gamorrean in the head with the gaffi stick. You have a Brawn of 3 and Melee of 1, rolling two green and one yellow due to your skill. You succeed solely with a success on the yellow die - With finesse, you duck under the Gamorrean's axe and plant the end of your stick in his teeth. Versus, you succeed solely with successes on the green dice - You pummel the Gamorrean repeatedly to the ground using only your strength. Or are you ignoring which scores came from what die and applying creative narrative regardless of where the scores came from?

Looking at which die provides the success and failures can help you with narrative, but I think it could significantly slow down play and/or get tedious after a while. On a related note, are GMs/players getting tired of detailed descriptive narratives after a play session or two? Is it becoming too repetitive and falling back on non-narrative descriptions such as, "you shoot him with the blaster again", or boring repetitive descriptions such as, "You squarely shoot him in the chest with the blaster", being used ad naseum?

We interpret pools at the end result, not as each aggregate die is rolled.

If narrative fatigue is working its way into your sessions then consider that you may be asking for checks and rolls too often out of combat. For example, there's no need to have everyone roll perception checks at every hallway intersection in a dungeon installation crawl stroll. I borrow a little of Burning Wheel's "Let it Ride" with Edge of the Empire, where a test result holds until the conditions or situation changes in such a way that a new roll is warranted. Otherwise we continue on abiding by the results of the previous check. And when the whole party may have rolled a check in d20 star wars, pair it down to one roller with assistance from the others for EotE. That way you aren't trying to determine how to apply a Triumph from two or three players at a time in the same situation.

Combat has the highest density of rolls in any game, and EotE is no different. Let the narrative stuff happen when it happens. Don't force it, or it'll be like having someone stand over your shoulder yelling "SING PRETTY!!!!" all the time ( I mean...I DO have the dulcid tones of an angel, but a kid can't perform all the time ). Don't feel bad about getting mechanical with applications of Advantage, Threat, Despair and Triumph in combat. Trying to make every blaster bolt and swing of a sword into something that the Ugnaught Spacebards will oink of for a thousand generations is going to wear anyone out. Despair and Triumph do lend themselves to a little special consideration, but deciding on a weapon jam or a crit isn't going to get anyone lynched for badwrongfun.

One thing I have done to keep myself and my players from slipping into mechanical adjustment complacency is to querry before results are applied with "anyone got anything cool to add before we dive in?" I'd say that about 1 in 4 combat rolls results in someone at the table seeing a pile of net Threat or Advantage and going "OH dude....I totally got something." And of course there are going to be some combat encounters that just have a certain je ne sais quoi about them, where narrative interpretations and inspiration flow easy.

Callidon is dropping some truth, right there.

I haven't run a ton of games just yet, but I've already started to find a groove regarding when to roll, and when to get flashy with the details of the outcome. It has a lot to do with which individual player is rolling. I've got players that I already know are going to look at the cheat sheet and spend their advantage by the book nearly every time- and some that are always reaching for the coolest narrative angle they can summon at a moment's notice. In either case, looking at each die in some determined order and attempting to ascribe a bit of any narrative to each die seems awful cumbersome.

I can easily see that technique leading to a potential narrative burn-out.

Final results, not going by each dice. It's the only way I know that would flow well.

I like Callidon's take. I've used all the alternatives Sturn described, in fact I've enjoyed using the particular dice to interpret a result if I'm feeling inspired. Basically it depends on what spirit moves me at the moment. I don't try to force it though, there's no faster way to kill the mood.

Which ever you prefer. There is no requirements to do it either way. The Narrative dice system is suppose to inspire ideas.

As the first example nails down the idea that if the black die was involved due to stand generates a fail. Then the GM/Player could certainly take inspiration that the sand storm was influential. However, if you don't want to don't. Choose what you think is better. I personally like the idea of the dice influencing the narrative direction, but that's me. If you want to just to use whatever the result is then do that

Callidon pretty much nailed it. The only thing I would add is divide up the narration duties within a roll. Let the players decide how the advantage and success are told and the GM gets creative with the threat and failure (or vice-versa with NPCs).

Also sometimes you need to gloss over or quickly resolve the less interesting dice results (e.g the 'everything cancels everything out' rolls) so that when you or your players get a great idea it stands out.

I occasionally use "uncancelled dice" to read the pool, and sometimes look at the total pool (cancelled and uncancelled)...

I've noticed my players have, subconsciously, taken to pulling greens in slight preference to yellows... They know that yellow is skill supported by ability.