aws s3 sync --existing

Could you implement a feature like RYNC's --existing so that you can tell AWS to "skip creating files that do not exist yet on the destination"

Or is there a way to do this already?

The use case is you have just synced from machine 1 up to S3 and you have a huge dataset in S3. Now you want to sync down from S3 to machine 2 so you're working with up to date files on machine 2. But you only want to update files that are already on machine 2, you don't want to copy the huge dataset in S3 down.

Make sense?