- Include object `id`/`key` in errors when objects not found in datafile
- Modifies invalid object `id`/`key` log messages to make them consistent

Include object `id`/`key` in errors when an object is not found makes them available to the user or the custom error handler.

One example of why this is useful is that the `key` of an experiment could be used within a custom error handler to fetch the details of the experiment. This would indicate whether the experiment has been paused (in which case the error could be ignored) or archived (in which case the code referencing the experiment should be removed from the application).

- Expose the object identifiers as error properties so that they are easier to use in error handlers.
- Encapsulate the error messages within the error objects to enforce consistency and to simplify initialization
- Use the messages of the error objects as log messages to enforce consistency and to simplify logging