Allow filtering by global and country site ranks, and report language / origin

The goal is to be able to do focused deep dives into the experience of users in specific countries. To do this we want to select all buckets containing reports that are:

• Top sites in that country
• Either originate in that country or were originally written in the local language (where there's a good mapping between language and region; that doesn't work well for primarily-English speaking countries)
• Not global top sites

Ideally each of these aspects should be individually selectable so that I can e.g. show all reports from France and/or in French either for all sites, or French top sites, or for French top sites excluding global top sites.

Use issue #147 as the starting point; no implementation files or tests are named in the report. Define how the independently selectable country rank, country origin, local-language, and global-rank filters should combine, then verify that the result supports the France/French examples and excludes global top sites when requested.