In sharded environment, a query like
{_id : {$in : [100, 200, 300, 400, 500, 600...]}}
needs to be re-written by mongos by first resolving each _id value to the shard based on shard key (provided the shard key can be deduced from the query) and then sending corresponding sub-qery to the right shard.
Example, if 100, 300, 500 belong to shard1 and 200, 400, 600 belong to shard2 MongoS needs to make exactly 2 queries:
{_id : {$in : [100, 300, 500]}}
to shard1
and
{_id : {$in : [200, 400, 600]}}
to shard2.