Write explains have two main code paths, each with their own translation logic. FindAndModify's clusterWriteWithoutShardKey code path requires extra work to cause "organically," the chunks of the bucket collection must exist across multiple shards. This might be achievable "organically" if enough data is put into the collection causing another shard to be created and migrated by the balancer. The test conditions can also be manually set up to get the same coverage.