Enhancement: Nested aggregation queries

[plcplc]

We can both sort and filter based on nested aggregations (i.e., through a path of relationships), but curiously, aggregate fields is only concerned with top-level aggregates.

This would seem like a nice thing to be able to do.

For instance, in the Chinook data model, it would allow selecting the total number of tracks that an artist have across all their albums (or, considering Paths have predicates, total number of tracks from albums published within the last 3 years).

[paf31]

For your example, you can compute the aggregate using a where clause and exists, right?

If we want to do the same for every track in a row set of tracks, then I think that potentially overlaps with group by functionality, and we could consider it in that context.