Split a large data frame in to a nested list and only keep meaningful non-grouping factors in R

Question

This is a complex follow up to this question Split a dataframe into a list of nested data frames and matrices I now split the original data frame with 2 grouping variables.