docs: fix stale TODO, hardware inventory, PD and Rocinante specs

docs/servers/PLAUSIBLEDENABILITY.md

@@ -31,7 +31,8 @@ Primary Docker host running all production workloads.
 RTX 2080 Ti used by:
 - Plex (hardware transcoding)
 - immich-ml (CUDA face recognition)
-- ollama (LLM inference)
+- Whisper (CUDA speech-to-text, port 8786)
+- Ollama light-tier inference (qwen2.5:14b and smaller)
 
 Recommended models for 11GB VRAM:
 | Model | VRAM | Best for |
@@ -45,9 +46,14 @@ Recommended models for 11GB VRAM:
 - Pangolin VPS + Newt for external services
 - SSH from Windows 11
 
-## NFS Exports to Other Servers
-- `data` share → PD at `/mnt/unraid/data`
-- `immich` share → PD at `/mnt/unraid/immich`
+## NFS Mounts (from Serenity)
+- Serenity `/mnt/user/data` → `/mnt/unraid/data` (media libraries)
+- Serenity `/mnt/user/immich` → `/mnt/unraid/immich` (Immich photo library)
+
+> **Note:** NFS mounts are not automatic at boot. Run after every TrueNAS restart:
+> ```bash
+> sudo bash /mnt/tank/docker/scripts/mount-unraid-nfs.sh
+> ```
 
 ## Known Issues
 - immich-ml intermittently unhealthy — investigate GPU/CUDA health
@@ -55,6 +61,6 @@ Recommended models for 11GB VRAM:
 - No off-site backup
 
 ## Pending
-- Deploy: dispatcharr, gotify, bytestash, kima-hub, n8n, qdrant, paperless-ngx
 - Add Pangolin resource + DNS for `ai.paccoco.com`
 - Install fresh editor
+- Set up private Gitea repo for .env file backups