Fast, approximate ES|QL queries for analyzing very large datasets

## Value proposition

Get near-instant insights from massive datasets by executing high-speed, approximate analytical queries. This allows them to make faster analysis in Discover and Dashboards without waiting for resource-intensive, exact computations.

## Expected outcome
Data analysts and SRE obtain fast approximate insights from very large datasets. Those include:
- An approximate aggregation value
- A range (confidence interval)
- A result reliability indication

### Example use cases:

- Discover: Fast Log Pattern Analysis when full-corpus categorization is slow on very large datasets.
- Visualizations: Create and edit metric visualizations over long time ranges that aggregate extensive historical data.
- Dashboards: Use dashboard actions on panels that include those long-range, aggregation-heavy graphs.