The AngryGM has some interesting suggestions for
tweaking action resolution in 5e
Not sure I'm on board with all of it but there's some good stuff in there and the general discussion is meaty.
Having finished the article, here are my opinions on the matter.
The Core Mechanic section is a little self-evident in my opinion, but there's nothing wrong with making sure your ducks are in a row before delving into deeper analysis.
The section on "Approach and Outcome, Consequence and Cost" is how my group has always played, but it's probably useful to those newer to the game. All in all, I agree in playing this way. I do think most of this would fall under the category of common sense, and therefore not really within the scope of RPG design. If you can't figure out for yourself that breaking down a door is LOUD, well...
The section of "No Skill Rolls" and "GM Decides the Rules to Use" are again already how we play. I wouldn't hesitate to call for a Charisma (Animal Handling) or Wisdom (Religion) check under appropriate circumstances. And I much prefer the player telling me what his character is doing than asking if he can roll a check (because, depending on the circumstances, I may rule that what they want doesn't require a check). Still, it's not a bad thing to remind DMs about, especially newer DMs who might be too rigid in their thinking.
The "Saving Throws are Weird" section is fine, and I actually really like the The Reaction Rule. I'd never actually considered freezing time the moment that a trap is set off and asking the player how they'd like their character to react. That's a cool idea. Although I would allow Bob's character a saving throw to catch the edge of the pit as he falls. Just because you didn't guess correctly (especially lacking any significant clues) shouldn't mean you are automatically screwed. The character could still react on instinct and catch the lip of the pit (depending on the roll I might require him to drop something he is holding in order to catch himself).
"Knowledge Checks are Weirder" is another thing we've been doing for a long time now. It just seems obvious to me. That said, I've seen DMs who didn't do things this way so it's a good thing to mention to those who might not have thought of it. To keep things simple, I don't set a DC. The main determination is simply whether the character is proficient (or doubly proficient) in a knowledge (it's not something I have to track as I can easily ask the player when it comes up).
Passive DC starting at an 8 is an interesting idea, but not one that I think I'll use. 10 is fine and mimics an opposed check while giving the "attacker" a slight advantage.
I disagree with him on "Teamwork and Group Checks", and I don't care for his house ruling in that area. In particular, I think that "Worst" Group checks is a terrible idea. That just means that certain approaches are barred entirely to the party. Have a clumsy fighter in full plate? You can never sneak anywhere as a group unless perhaps you can convince him to take off his armor and attune to Gloves of Dexterity (or whatever). The current group check mechanic allows the sneakier characters to make up for the less sneaky characters, which enables cool scenes (the rogue's player can narrate how he creates a distraction to cover the fighter's noisy approach) and having all the players roll a check is only minimally more time consuming than having a single player do so.
Overall though, a well considered article.
EDIT:
Expanding on the Reaction Rule, I would probably give Alice's character advantage on her saving throw, but have Bob make his normally. IMO, it's better to reward a lucky guess than to penalize an unlucky one (plus, there are likely to be a very limited number of "right" answers and a virtually infinite number of "wrong" answers, so this approach is much more fair). If the DM foreshadowed the type of trap (gave the players several clues) then a harsher method of resolution might be appropriate.