Release memory eagerly in ordering and distinct accumulators

[pettyjamesm]

Description

Releases PagesIndex from Ordering(Grouped)Accumulator and MarkDistinctHash from Distinct(Grouped)Accumulator instances after prepareFinal() or evaluateFinal() is called.

This helps mitigate an issue exposed by #25879 - which was that HashAggreationOperator wouldn't track any changes in memory consumption after output started. However, ordered grouped aggregations significantly increase their memory consumption at the beginning of output as a result of transferring the contents of their PagesIndex into their underlying accumulators.

Release notes

( ) This is not user-visible or is docs only, and no release notes are required.
( ) Release notes are required. Please propose a release note for me.
(x) Release notes are required, with the following suggested text:

## General
* Reduce memory required for distinct and ordered grouped aggregations ({issue}`26276`)
* Fix memory tracking for ordered grouped aggregations ({issue}`26276`)