Migrate away from using "myna_data.yaml" for sending metadata to cases

[gknapp1]

Currently the "myna_data.yaml" file is generated for each case. While each case then has access to its own copy of the metadata, each case is not entirely self-contained as any file-based metadata still lives in myna_resources in the root Myna project directory. This creates a downside where the information is segmented between cases and the main Myna input file (e.g., "input.yaml") and additionally each application needs to handle parsing the myna_data.yaml file.

It would be more consistent to have each MynaApp-derived application use a base class function to obtain the metadata that it needs. Something like MynaApp.get_metadata(metadata_name) with "metadata_name" corresponding to the names in myna.core.metadata.data_class_lookup().