Might be nice to have some functionality that automatically tries to split the translation codebase into a certain number of equal-sized sections to facilitate translating as a team.

Something like

translate_package(team_size = 4, team_id = 1)
translate_package(team_size = 4, team_id = 2)
...

Maybe do something like (1) "block-randomization" to try and make sure each team_id gets assigned messages of roughly equal complexity (e.g. essentially frank(-nchar(msgid)) %% team_size) (2) try and match each team_id to messages totalling roughly sum(nchar(msgid))/team_size (3) block by files to try and leverage some efficiency from translating similar messages by the same person?