"Each frame is composed of 1 to 4 of the following colorspace planes: Y, Cb, Cr, Alpha"
This is true for YCbCr only

Right. I also think this section needs to be changed to reflect the hierarchy of frame>slice>plane, as currently this paragraph only refers to frame and plane, and then slice arrives later in the documentation without a good introduction.

I'm reviewing this again and have some questions:

• When the colorspace is RCT, are planes used? When I encode rgb24 with ffv1.v3 I only see one plane, but may be misinterpreting this.
• Re: "Each frame is composed of 1 to 4 ... planes". I'd like to be more specific. From this reading it seems like 2 planes is possible but that Cb should follow Y; however, are ya colorspaces possible (grayscale with alpha)?
• Is a 2 plane encoding possible?
• What is the meaning of line in this paragraph? The line count appears effected by the colorspace and possibly the chroma subsampling, but I'm not clear about this.
• Re: "since it is most likely that the RCT will immediately be converted to RGB during decoding" what other conversion is possible?

i dont think pull requests are the right place for such questions, theres a IETF and a ffmpeg mailing list.
Here in the pull req basically only 2 people will see it

I moved the questions to an email here: https://mailarchive.ietf.org/arch/msg/cellar/nw5nwEmfnEEBu88QSOyJbsOQu5E. It appears the commits of this PR are already in the repository so I am closing the PR.