Not super familiar with the v3 codebase yet but from a quick look, this feature looks the same as on v2. Maybe add a comment on the feature request you linked above with a bit more information on how you would want to use this feature (if you can think of your ideal API)? Just a thought