[Lens] [EsAggs][Meta] Optimize esaggs requests when possible

[dej611]

Describe the feature:

Lens is a generic analysis tool and leverages the esaggs infrastructure to query data from Elastisearch: in general the approach has been so far to not make specific assumptions based on fields or index metadata to build the aggs request to send over the wire.

I think as we are approaching enough coverage of supported aggs in the editor it makes sense to operate on the optimization side and leverage some metadata to reduce the query time when possible.
In the past some work has been already done to reduce the number of aggs (as for duplicate percentiles or formula duplicate operations) but this can be extended also to specific cases where ES offers speed-ups for specific cases.

Issue breakdown

Preview Give feedback

1. [Lens] improve percentile agg optimizations #145303
   Feature:Lens Team:Visualizations backport:skip release_note:skip v8.7.0 +5
2. [Lens] optimize duplicate formula functions #140859
   Feature:Lens Team:Visualizations backport:skip release_note:skip v8.6.0 +5
3. [Lens] Optimize redundant formula function aggregations #135265
   Feature:Lens Team:Visualizations impact:medium performance +4
4. [lens] [optimization] use es rate aggregation when possible #152537
   Feature:Lens Team:Visualizations blocked enhancement epic impact:medium +6
   
5. [Lens] Use rate agg behind the scenes for counter rates on mode time_series data views #130714
   Feature:Lens Team:Visualizations enhancement impact:high +4
   
6. [Lens] Optimize advanced filter by queries #153628
   Feature:Aggregations Feature:Lens Team:Visualizations enhancement impact:low performance +6

[elasticmachine]

Pinging @elastic/kibana-visualizations @elastic/kibana-visualizations-external (Team:Visualizations)