Improve Parkinson research pages with patient-specific evidence

[PriyanshiYadav007]

Summary

Reorganizes both Parkinson project pages around actual patient evidence, following
Prof’s September feedback.

• Paper 1 compares two patient histories, showing measured sources, missingness,
  forecasts and the limits on trusting those forecasts.
• Paper 2 presents four selectable cases covering opposite laterality, near-symmetry
  and imaging–clinical disagreement.
• Separates measured state, held-out forecasts and observed follow-up through an
  interactive replay.
• Keeps uncertainty, unfavorable results and clinical limitations explicit.
• Updates scientific figures and connects both pages to earlier CVC research and
  AI4PD.

No manuscript files, unrelated project pages, participant identifiers or raw cohort
tables are included.

Verification

• Seven data and link tests passed.
• Both pages passed desktop/mobile browser checks.
• Scientific records rebuilt from frozen analysis outputs.
• Full Gatsby build was not run; the standalone static pages were tested directly.