First of all, an apology for taking so long to respond to your email. The information you have requested, in spite of it seeming quite obvious, is it hard to find and this is due to lack of documentation by the team of GlusterFS' Development Area.

A point to take into consideration is that our implementation of the  filesystem is based on three different projects that are related as follows:
 * btrs in the bottom layer
 * glusterfs in the middle
 * A reimplementation of Swift at the top

The information is somewhat scattered.

In summary:
- Is said that s3 and openstack swift are compatible but does not mention to what extent.
- Some features of openstcak as the ring builder is not yet implemented in GlusterFS.
- Btrfs already provides copy-on-write, as well as snapshots of the file system and native compression but this is not exposed to the GlusterFS layer.
- The communication between servers is not encrypted, which is why we use the private network to handle the traffic generated by GlusterFS.

Another important point is that the Gluster layer that provides support for S3 (UFO) has not been updated since June last year, and discovered the reason just an hour ago. It seems that the developer of the project has stopped to work on UFO and is probably working on the plugin for openstack-gluster. I will try to contact him to get more information about it.