The collection is targeted to have 1 billion documents and the subdocument field is targeted to have 100 documents.

This is equivalent to a collection of 100 billion documents.

Besides indexing the subdocument field, what are other scaling tactics that I can take?