data request csv vs yaml file

@mvertens i see some value in starting from the csv file instead of a yaml file
as you demonstrated this morning. It would make clear which fields each model is providing and which are not available or still need development.

But in that case why have a yaml file? Why not just use the csv file directly?

And do we need to have separate csv files for cesm and noresm or can we just have separate columns in the same file?

If we are to embrace this approach I would need a script to fill in the csv fields based on the current cesm_to_cmip7.yaml file.