Something to let go would be that each pillar of play has differing pre-defined levels of granularity instead of the time spent on it being based on how important it is to the players.
So if (uncertain) activity X is not of much import but we do have some moderate level of care about the output, solve it with a die roll (or whatever). If it's more important, have it mechanically heavier (though don't weight it more towards the average with "more of the same type of roll"). If it's of great importance to the players, then zoom down to a high level of granularity that's going to take some time to resolve.
To use some high fantasy examples, a wandering monster, haggling for expensive items, navigating to the next port or tavern brawl might be dealt with with a die roll, because they are of low import to the player. And enough levels of response that it's not "you succeed/failed to navigate" but also "it took you three extra days" or "some other dwarf out-drank you and you came in second" or whatever. Note that inherently means that we aren't artificially increasing stakes by taking a scene that the players don't care about and out of nowhere assigning arbitrarily high stakes like possible character death in a system designed for resource attrition.
On the other hand, mayhaps being able to flow the tracks of the kidnapper is quite important, and especially do they arrive before or after the "extraneous" kidnappees are disposed of. That might be zoomed in, with more rolls and some meaningful choices about approach. Same thing for a moderate skirmish, or convincing the sheriff that the party didn't steal the dowry chest.
And a final level that zooms all the way in, only for scenes that are the most important to the players. They should involve everyone, and have meaningful choices made along the way that inform the result. This is probably the only level that might get down to task-resolution granularity. This is the big "convince the duchess her husband is betraying her and the duchy" or "stop the ritual turning people into abominations" or whatever. These come around only once or thrice in an adventure, a few more times that that in an arc.
The idea that all of activity X is zoomed into THIS level and all of activity Y is zoomed out to THAT level, without being able to focus session time on what is actually important to the players, is an relic that does not serve the table.