Improve OWID extraction

[rapsoj]

Depends on previous PR. Improvements to OWID extraction.
This pull request significantly enhances the trend analysis capabilities in the OWID custom scraper. The changes introduce detailed statistical summaries and model fitting (linear and exponential) to the trend output, improving the quality and interpretability of time series insights for epidemiological data, based on Karla's findings.

Trend analysis and summary improvements (_owid_common.py):

• Added statistical summary and model fitting utilities: functions for R² calculation, linear and exponential fit, and formatted statistics summaries to support richer trend analysis. [1] [2]
• Enhanced the _trend_table_html function to include:
  • 7- and 30-row deltas, increment statistics, and model-fit results for both cumulative and incident (new cases) data.
  • Clearer, more detailed HTML output for target entities.

[smodee]

Thanks @rapsoj! Extracted and merged: the OWID trend/model-fit work as #72, and the PAHO scraper as part of #74. Leaving this open for you to close when ready.

[smodee]

Closing with @rapsoj's OK — superseded by #72 (OWID trend/model-fit) and #74 (PAHO scraper). Thanks Jess!