Wrong lead time in Efficiency dashboards (gauge visualization)

[canasdiaz]

When showing the lead time in the GitHub efficiency dashboards, we either have the wrong documentation or metrics. The lead time presented in the gauge is only using the items created in the selected time frame. This needs to be corrected. For example, if we select the last 6 months, the lead time would not include anything about a ticket opened for three years and recently closed.

How can this be solved? I see two options:

1. we update the documentation both in htps://chaoss.github.io/grimoirelab-sigils/ and in the panel itself to include this limitation
2. we do this calculation with an index pattern that uses the closing date as the default time field instead of one using the creation date

[fioddor]

At first glance, the second one makes more sense to me as the usual case: How old is the stuff we're closing in the time period.

But the first one could make sense under certain circumstances: Imagine we want to accelerate. What can we do? We can change templates or tags or automate checks. These actions affect the new stuff, so we'd like to exclude stuff created before.

And I can expect the user to expect that by default the time picker filters the creation date of the elements for most visualizations.
They can extend the time-picker and apply a filter to the closing date.

Regardless of the solution we chose, the documentation needs to explicitly disambiguate.