Aah, from Maximum file size on ext4 - #3 by SteavedHams
One reason for this is that FDB shuffles data around a good bit in response to writes, and any time the cluster relocates a data range from one Redwood instance to another it is effectively “compacted” as the destination will have a condensed low-slack (or even negative slack) subtree of that data and the source will delete the subtree completely which frees all of its blocks.
So if a shard is moved the entire subtree moves, so disk space can be (and is) reclaimed on the origin process?