Allow extract validate to check a subset of data files

It may be valuable to allow extract validate to check a subset of input data files (e.g. the first 10 years) if the amount of data files is otherwise too large (e.g. files spanning longer than 50 years) . This would improve efficiency and allow for some level of validation even for the largest of data sets.

Start by locating the `extract validate` command entry point and reading how it currently selects and checks input data files. Clarify the intended subset selector, such as a year range or file count, and define done as validating only that selected subset while preserving current full-dataset validation.

Written by the indexing model from the issue text.