drivers: video: improve the format description docs

@verbatim shows the leading comment '*' fence in the output.
@code{.unparsed} allows extracting the text only.
In some formats, the pixels are also not immediately packed into fixed
number of bytes, like YUYV due to chroma subsampling. Disambiguate.
Add numbers marks on top to help identify the individual pixels.

Signed-off-by: Josuah Demangeon <me@josuah.net>