I think I may have misinterpreted the docs and mixed things up.
From what I understand, the SES admin is only needed if you want to run a SES Ceph cluster as a standalone storage back-end. If you already have a SUSE OpenStack Cloud running with the respective admin node you won't need another admin node, it should be sufficient to install the basic ceph packages on the SOC admin to also use it as the ceph admin node, if required. That's what I successfully set up in a virtual lab environment, that way I was able to use an existing (also virtual) ceph cluster for glance and cinder. This would have also worked without installing the ceph packages on admin node and directly work on the ceph nodes, but I like the ceph-deploy tool and wanted to use it remotely.
This makes a lot more sense to me, but please correct me if I'm wrong! ;-)