Add interpretation guidance for metrics and suggest_resolution output

[natalie-23-gill]

Problem

Users get a ranked table from suggest_resolution() but the docs don't explain how to act on the results. Key questions left unanswered:

• What does a flat rank curve mean vs a clear winner?
• When should you use "local_optima" over "rank"?
• What Hellinger distance values indicate poor reproducibility?
• How should the silhouette score distribution inform the final decision?

Proposal

Add an "Interpreting Results" section to the vignette covering:

• How to read the rank plots and what different curve shapes mean
• Decision framework for choosing between "rank" and "local_optima" methods
• Rules of thumb for metric values (e.g., Hellinger thresholds, silhouette score ranges)
• Examples of ambiguous results and how to handle them