Implement DIC and other model evaluation metrics?

Following on from this thread:

https://forum.greta-stats.org/t/deviance-information-criterion-dic-in-greta/161/9

And

https://forum.greta-stats.org/t/model-selection-criteria/319

Begin with the linked forum threads on DIC and model selection criteria to understand the requested metrics and intended behavior. The issue does not name files, tests, or an entry point; done would require defining the metric scope and implementing the agreed model evaluation functionality.

Written by the indexing model from the issue text.