Add documentation on how to get/generate CSVs used as input for Workbench delete/update/add_media/etc tasks

[mjordan]

Sources include:

• Views
• Workbench export output
• Workbench's rollback CSVs
• Workbench's CSV ID to node ID map

This might be a good opportunity to solicit contributions / local workflows from users!