Description
Current parameters and attribute files can vary both along the hru and gru dimension which makes it at times difficult to process or subset them with general-use utilities. For example, @andywood comments that "one can't really to a simple ncrcat to paste a domain back together, or subset files on the hru dim without getting all the grus"
Splitting them out into files that only vary along either dimension is not entirely straightforward because the order in which the files are processed matters (https://summa.readthedocs.io/en/latest/input_output/SUMMA_input/#attribute-and-parameter-files).
Revisiting file organization and processing order to simplify this may be worthwhile. This would break existing setups, so should perhaps only be done as part of a major version upgrade.