[QuickAccent] Support custom ordering of character sets

### Description of the new feature / enhancement

Character sets used to be ordered by their placement in the source code, which was essentially invisible to users and not based on either the language name, our designator, or language ISO code. Recently, this has been updated so ordering is predictable and based on the alphabetical order of the name. However, this is not ideal for some users because they may have become used to the previous ordering and this update has broken muscle memory.

This feature would allow the user to order their chosen character sets as they see fit, and potentially allow them to test out the changes in real time.

It is not intended to restore the old source code ordering, i.e. by default the new ordering will remain in place for predictability.

### Scenario when this would be used?

The new feature would be used when the user finds that the ordering of characters for their chosen character sets is not to their liking and they would prefer to control the priority explicitly.

### Supporting information

Please see #48488 for further information of user-facing concerns about the recent change to character set ordering.

Start by reading the QuickAccent implementation and the discussion linked in #48488 to understand the current character-set ordering and the user-facing concerns. Define how users would configure and preview a custom order while preserving the predictable default ordering; completion should include a clear interaction design and validation of the proposed behavior.

Written by the indexing model from the issue text.