Legibility is the closest correlate of success

In life we usually have somewhat legible goals with somewhat legible paths to get to them—studying and building a portfolio to get jobs, or passing exams to become qualified for particular types of work. But many of the most valuable things that we want (agency, energy, happiness, good companionship) have no working consensus paths or definitions.

Legibility is important because it’s one of very few factors that actually help with meta-level pathfinding, rather than just walking a known path more effectively & efficiently. It helps us search and reduce for the specific set of (presently known) variables that gives the best leverage in moving towards a goal; without it there are enormous amounts of capacity and optimization left on the table.

When something is illegible it's hard to make progress on them because failure provides very little improvement (i.e. path-correction) value and we can get stuck in the local landscape. If we think that the problem is in not grinding/working/efforting something enough (though sometimes these are the bottlenecks!) rather than having a poor approach or hyperparameters, we can get trapped in myopically focusing on single-digit %/day changes rather than available 10x-1000x leaps.

Caveat: probably should be cautious of applying this dogmatically to open-ended questions, especially research, where you probably need to go through many loops of making different specific paths go from illegible to legible (to figure out if they’re worth further pursuing), and tunnel-visioning on making one early trailhead you find super legible is likely suboptimal.