Almost everything that the UI does is possible to do via API:
As for getting any features to the feature, that’s not going to happen. We’re not actively investing in that feature currently and the recommended way to implement split testing is using Edge Functions.