Schema design: Provide suggestions to accommodate or correct poor design

Provide help for working with or correcting scenarios (common problems?) like the following:

New user feedback!
Page: https://docs.influxdata.com/influxdb/cloud/write-data/best-practices/schema-design/
Feedback: Hello. This article does not offer guidance how to handle case when tag has many values. Real-world example: I have a test system which produces 100 000 points (field values) for one test run. The number of test runs itself is in order of 1 000 000. That leads to "testid" tag with 1 000 000 tag values, with each tag used to tie together series of 100 000 points for each test. This is discouraged, obviously, but how can I count, say, individual average for each test run and put that into bins, if I cannot filter on "testid" tag?

Start with the linked schema-design article and the user feedback in this issue. Determine what guidance is needed for a testid tag with 1,000,000 values and how users can calculate per-test averages and group them into bins. Done means the documentation addresses this scenario with an actionable recommendation.

Written by the indexing model from the issue text.