Improve aggregation docs for Redis HLL events

The following discussion from !48457 (merged) should be addressed:

• @mikolaj_wawrzyniak started a discussion: (+1 comment)

  Thought (non-blocking): I wonder if instead of simply encouraging weekly aggregation, we should help PM to understand differences between daily and weekly aggregations, and how weekly should be sufficient for the most of cases.

  Form what I picked up in code and in the documentation of usage ping https://docs.gitlab.com/ee/development/product_analytics/usage_ping.html#known-events-in-usage-data-payload

  • #{event_name}_weekly: Data for 7 days for daily aggregation events and data for the last complete week for weekly aggregation events.
  • #{event_name}_monthly: Data for 28 days for daily aggregation events and data for the last 4 complete weeks for weekly aggregation events.

  In context of business analysis, the only difference that is between both aggregations is how weekly and monthly time spans are selected for counting unique events, but selecting daily aggregation does not bring more detailed or fine grained data.