I'm getting some weird behavior using the RawBatchCursor and I'm not sure if its normal or not. For instance if I run an aggregate_raw_batches and expect to get 10,000 documents back and run `list(my_cursor.batch_size(1))` I get a list of 5,000 elements with each element being 2 documents concatenated together. Then if I insert this I get 5,000 documents inserted (first document from each pair) but I would expect 10,000.

What I would like to do is be able to run `insert_many` on the resulting RawBatchCursor of an aggregate_raw_batches and have it upload all docs instead of the first one in a concatenated string of bytes without any decoding. Could you advise on the best way to accomplish this?