fix(paragraph-ids): deduplicate reused glossary IDs

[PLeVasseur]

Summary

• Deduplicate two reused :dp: IDs inherited from mainline by updating glossary entries only.
• Keep canonical IDs in src/types-and-traits.rst unchanged and re-ID the duplicate glossary entries.
• Reduce duplicate paragraph IDs across src/ from 2 to 0.

Before

• fls_t4yeovFm83Wo appeared in:
  • src/types-and-traits.rst:652
  • src/glossary.rst:1807
• fls_I9JaKZelMiby appeared in:
  • src/types-and-traits.rst:2742
  • src/glossary.rst:3930

After

• Canonical chapter IDs unchanged:
  • src/types-and-traits.rst:652 remains fls_t4yeovFm83Wo
  • src/types-and-traits.rst:2742 remains fls_I9JaKZelMiby
• Glossary duplicates re-IDed:
  • src/glossary.rst:1807 -> fls_kqdvWGi9cglm
  • src/glossary.rst:3930 -> fls_H5vkbMFvzrFs

Reference alignment

• No language-rule semantic changes; this is paragraph-ID hygiene to restore uniqueness and keep ID-keyed tooling deterministic.

Testing

• ./make.py
• Duplicate :dp: inventory before/after (before: 2 duplicates, after: 0 duplicates)