Document playbook run requirement for old installs

[eloquence]

During a recent install, we encountered ETag errors from the API, which indicates that the Ansible playbook that applied a relevant change had not been run. The issue was resolved by running an up-to-date Ansible playbook.

We should document any playbook run requirements needed for successful SecureDrop Workstation use with a long-running server.

[redshiftzero]

Orgs need to have had a playbook run on 0.13.0 or later to include these changes - e.g. orgs that upgraded to Xenial (running the playbooks) right after 0.12.0 was released will need to run the playbooks once again.