One of the worst part of d&d is skills is the pass/fail, there isn't really a degree of success, which is the exact opposite to combat (where degree of success is hit points damage)
If you change that perception a little (and it's kind of embedded in optional rules) think about it from a level of success, if you like somewhere between no roll and passive.
Basically a passive check gets you enough - need to climb a wall with a ladder beside it - that's a DC10... Average person suceeds, taking their time and climbing the ladder one step at a time ... You have a +5 dex and proficiency, sure up you go in half the time, only takes a bonus action ... Want yo take a risk and go for it being a free action v dc20 ... Roll away
Same with clues or knowledge.... Passive you know trolls regenerate ... Int +5 with proficiency, but in this part of the world it's acid and not fire that stops it... But what was that thing they take double damage from... Want to roll?
Etc....
Low passive - basic; high passive advantage over a basic score; take a roll, risk reward payoff
Agreed, which is why I incorporate more options than just pass/fail. There are a lot of ways to do it, having potential consequences if you fail by more than 5, or more than 10 are common choices. But I like the concept of Take 10/Take 20 in that it takes into account that a person capable of succeeding at something eventually will, given that they have enough time to keep trying.
Consider shooting a basketball. The old approach of if you fail once, you can't try again until you go up a level makes no sense. That assumes that until you improve your skill you can't possibly succeed (even if none of the effects of going up a level directly impact the skill in question). That just doesn't make sense to me. You just keep trying until you finally succeed.
The reality, if there is no danger at hand - you found a chest with a lock, and you can work on it as long as needed until you pick it - there really isn't any point in making a skill check. And if there is a danger, like picking the lock to get through the door before the guard returns, you could just keep rolling skill checks. But constantly re-rolling a pass/fail check is also kind of boring, although a valid option.
By tying failure in that case to time, you failed by 4, therefore it's going to take 4 rounds to succeed (the player just knows it will take longer), then the act of picking the lock continues while the rest of the characters find a way to give them enough time to succeed. They know it will take a few moments, and they have to provide the time needed.
This works well for most situations. Trying to remember that specific fact? "I know who that is, it's on the tip of my tongue...it'll come to me..." are situations that play very well.
It's all part of my general philosophy that the rules should provide us methods to adjudicate something specific, based on the skill of those involved, but allow for that bit of randomness without getting in the way of what's happening to the characters. Every time you roll a die, you are pulled out of what's going on in the game world, essentially "pausing" the action to determine what happens. Limiting the skill check to a single roll (until the situation changes significantly) meets those goals. The skill of the PC is taken into account, randomness is added, and the results can easily be determined.
Passive checks work within that same framework, there is less randomness with only one side making a roll, but that already has precedent with the attack roll vs a fixed AC. More importantly, I think it's a good representation of a character who isn't actively attempting something. For example, a merchant is attempting to deceive the PC. Unless the PC has reason to suspect the merchant from being dishonest, there's little reason for a skill check. A passive check is reasonable, because it's what represents the character's natural observation, what might tip them off that something's not quite right. You don't want the players to suspect every merchant, that's not realistic or fun after a while. Passive checks fill that gap when you as the DM knows something the players don't.
It's also a good tool for groups that don't like the DM to roll secretly. In those cases, they can adjudicate a lot of actions fairly without requiring the players to make a die roll that would tip them off that something might be up, and how well they have done.
Ilbranteloth