Test and improve .model.material

[macflo8]

This is a tracking issue for follow-up items from #188 and #189 . Create 1 or more PRs and (if necessary to keep discussion organized) issues to address:

• Add more tests to raise coverage to 100% (@macflo8, @GamzeUnlu95) (Extend MESSAGEix-Material tests #240)
• Refactor MESSAGEix-Materials reporting #192
• Develop better input data system for MESSAGEix-Materials #193
• Extend documentation of materials code (@macflo8, @GamzeUnlu95)
• Rebase https://github.com/iiasa/message-ix-models/tree/migrate-materials and decide how to move forward with Add MESSAGEix-Materials module #130 (Edit: Add MESSAGEix-Materials module #130 will not be merged, but is archived to remain as a reference to past projects)

[khaeru]

Another item to add: packaging

• Looking at this workflow run, we see a number of .xlsx files in /data/material/ are included in the build .whl and .tar.gz distributions.
• For the .water module and others, we have taken the approach of excluding these files from the distribution, and providing a function/CLI command to fetch these files from GitHub. See Miscellaneous enhancements for 2023-W44 #137 (mixed with many other changes, sorry). See also Data, metadata, and configuration in the docs.
• With that approach, the installation of message-ix-models remains small, and then users can fetch the data that they want to use. We don't force e.g. .water-focused users to download .material data, or vice versa.
• I would like to do this before we make a new release of message-ix-models that includes these materials .xlsx files.
• This is more or less orthogonal to Develop better input data system for MESSAGEix-Materials #193, i.e. it applies whether the data files are CSV or Excel or any other format.