[jc-macdonald]

The introductions and worked guides used promotional comparisons, universal performance claims and statements that could imply a Pareto front or a second simulation phase establishes scientific validity. Replace these with concrete descriptions and results scoped to the evaluated configurations.

Simplify the README and documentation overview. Clarify Pareto dominance, distinguish predictive scores from calibration, remove unsupported timing and optimizer comparisons, and describe the monitoring model as a simplified observation model. Keep equations, algorithms and runnable examples unchanged.

Validation: just format, just check, just ci, just doctest and uv run mkdocs build --strict passed. The full suite has 883 passing tests with 95.25% coverage. No runtime, dependency or version change.