This ticket should figure out and extend the tools/backup_analysis.py script to analyze backup differences based on collections and indexes. This ticket should:

• Determine if a new argument is needed to do or don't do this analysis.
• Use our knowledge of MongoDB's naming convention to track how many bytes changed in collection*.wt tables and index*.wt tables. Tracking should be done on categories of collections, indexes and "other" tables.
• Find out if there is any guaranteed way to know what index tables go with which collection based on naming. I don't think there is, but if we could determine that, it would be much more useful.
• Determine what information is relevant or interesting to report about collections and indexes.
• At a minimum how many bytes are copied for collections versus indexes would be useful to report.