What does it mean that schemas have resolution limits?

Every schema captures some details and loses others — resolution is a design choice.

Example: A project manager's schema for 'team health' tracks velocity, sprint completion, and standup attendance. It captures throughput perfectly. But it cannot see that the senior engineer is burned out, the junior engineer is afraid to ask questions, and the designer has mentally quit. The schema has resolution for productivity but not for morale. The data looks green while the team is falling apart.

Try this: Choose one schema you actively use — your model for what makes a good hire, a healthy relationship, a successful project, or a productive day. Write down the 3-5 signals that schema tracks. Now write down what that schema cannot see. What dimensions of reality does it have zero resolution for? You'll likely find that the blind spots are where your recent surprises came from.