So one of the shortcomings of the D20 system that DnD and other similair TTRPG's run is the binary nature of a challenge, you either succeed or fail. Other systems allow far more flexibility, for instance the roll and keep mechanic allows the possibility of anything from a catastrophic failure, to just failing, to a poor success to a fantastic success. This means you can give a lot more scope to results. 


You vault over the sail / you fail to vault over the sail (DnD) turns into 


No success - You catch you foot as you vault over the sail, your now hanging upside down 
Success but not enough to make the Target - You vault over the sail landing heavily, you lose your momentum, you lose the rest of your movement
Success hitting the target - You vault over the sail, it isn't pretty or heroic, but you do it, carry on as normal
Success with just a few more over the target - You vault over the sail landing lightly on your feet right in front of the enemy, and you look fabulous doing it 
Big success - You vault effortlessly over the sail, springing off the mast, you gain an extra 5 feet of movement this turn from the extra distance you gain. 

In a roll and keep approach

So I have started using scaled DC's a bit more for non combat, setting a range of DC's in my head that provide varying degrees of success/failure. 

Just wondering if this is an approach anyone else takes? 

The concept IS used in 5e.

ex: Roll a DC XX saving throw.  If you fail by 5 or more, something bad happens.

As a DM, you can embellish this all you want.  I do.

Since I am usually improvising these things, I set a "soft" DC in my head and decide the outcome based on the die roll's relation to that. I am wary of the "Big Success" category though because of the ways a few classes can stack skills so high that they sail way over the DC on a regular basis and may begin to expect to get extra bonuses on a regular basis. Generally a version of "you look fabulous doing it" is enough to make the player happy.

For what it's worth, the binary outcome only determines success or failure. It is still up to the DM to determine how the success or failure plays out. This is where 5e encourages (but doesn't insist) that the how should be related to degree of success/failure.

I also like the "success but with complications" concept. Often when they barely miss the soft DC, I'll let them succeed but there's a drawback. Maybe they vaulted over the sail and got to where they wanted to be, but they're a bit off balance and grant advantage to the next attack. I have found that players will happily accept other penalties - even harsh ones - as long as I let their plan work.

I'm a big fan of Fail Fowards.

Let me explain.   I played in a game where the DM had the party climbing a mountain, crossing precarious trails.  At one point he made the party of 6 roll DEX saves when a portion of the trail started to collapse.  Out of the 6 players, inevitably someone failed.  

Then the DM allowed someone to react and try to help, and made that a skill check too.  And that failed.

We continued to have a run of bad luck and no matter what we tried to do to undo the failed skill checks, we continued to fail.  And the prepared outcome was: one or more characters were going to fall thousands of feet to their death on a failed check.

But the DM didn't have the guts to go through with it and tell any player, "Sorry, but you failed that useless skill check that I put there for no particular good reason and now your character is dead."  So he kept giving freebie skill checks to save the situation.  It ruined immersion and wasted over an hour of gameplay.

Instead, he should have Failed Forwards:

"You fail your dexterity saving throw and slide down the face of the mountain until your hand gets wedged in a fissure, stopping your descent.  You find yourself precariously hanging on the side of a mountain."

Followed by a consequence:

• You take 1d6 bludgeoning damage, or
• The stress of knowing you're one handhold away from dying is tremendous.  After this ordeal, your character develops a fear of heights, or
• The effort of climbing back up to the path is exhausting.  You gain 1 level of exhaustion, or
• In the tumble, your backpack gets torn and you  lose half your rations (or something else.  I'd roll for a random outcome if potentially losing something else)

The intent of a skill check should be to consume resources, ie: HP, Exhaustion, equipment.  Or add depth to a character, a la "you now have a fear of heights".

I apply this even in a battle.  In the example of vaulting over a sail, I would not ruin a player's turn by saying they can't reach the target so their turn is done.  I'd say you fumble badly and land prone at your target's feet.  You're out of movement so you can only attack at disadvantage (for being prone) [and melee attacks will be at advantage against you].  It's a bad outcome for a bad saving throw, but at least they get to do what they intended to do, which is attack a target.

Absolutely. Even in some stat blocks there are examples of this. Drow is one that pops into memory. I pull this one out when there is something to be gained or lost, yet there is no real reason that the PC should outright fail to do the "thing". It's also a good way to add complications to a normally mundane travel montage, or say a PC wants to spend some time pickpocketing for their downtime. There's a certain level of competence that the PCs should be able to attain that allows for success, but maybe at a cost if they aren't paying strict attention. 

Helping an "old rug merchant" repack his overturned cart is a kind gesture, but failing that Perception check to notice the Rugs were not normal is almost too good to pass up.