Optimizing Mongo data model and query for large embedded collections

Question

I would like to optimize a document structure/query for a data access requirement below.

Batch data model

{
    "id": "