Having a short list of specific use cases / scenarios might be helpful as well. And a description of the desired API and features would be awesome (e.g.: "Have a stereographic/orthographic camera pointing at ??? at a distance/angle determined by ???, let me feed in some data in the form of ???, I need to use colors/palettes for ???, I'd like to have a grid/axes/labels" etc). That list could then be expanded into "How do I X in Y?", perhaps even with boilerplate code and/or helpers.