BQ finops -> BQ Scans by Table Type (evaluate your scans efficiency) -> the values in the graph are from different part of the data definition - one value is on the very higher level ‘view’ (this is not a table type) -> can use partition / cluster in the underlying query. other is specific table type - ‘clustered’. and others are types of partitions - ‘tableRangePartition’ / ‘customPartition’ / ‘ingestionTimePartition’.
If we show values from the same group and add a drill-down option, customers will get better insights on the query usage.
I also think that by the end, the graph trying to tell the customer whether his queries usage in the project is "good" or "bad", and this is the main insight he can take - so we can add a summary like: you are 90% good or a bar with a mark between good and bad.
I am sure that the academic routine is not what modern students need. Try this out, and you'll see that delegating assignments is a good idea.