Add confidence intervals to SCI scores

[PawanRamaMali]

Summary

The SCI is currently a point estimate. Users need uncertainty quantification to make informed go/no-go decisions.

Tasks

• Implement bootstrap-based confidence intervals for pillar scores and composite SCI
• Add compute_sci(..., ci = 0.95) parameter that returns lower/upper bounds
• Propagate indicator-level uncertainty through the scoring pipeline
• Visualize confidence intervals in the summary output (e.g., SCI = 82 [76, 88])
• Document interpretation in a vignette section

Why

A point estimate of SCI = 72 could mean "solidly conditional" or "borderline ready depending on measurement noise." Decision-makers need to know the difference.