My environment is far from simple.
Three Intel NUCs:
HAOS
Plex
Ubuntu (for general Linux programs)
and two Raspberry Pi’s:
MQTT broker
HAOS Sandbox (for experiments)
I just find it so much easier to manage individual servers.
(Why an MQTT Broker? I started out in IOT using Node Red on the Raspberry Pi and the IP address of the MQTT Broker was hard-coded into my devices, so I just kept the separate broker instead of reprogramming all the devices.)
That is a wire mess! 
But it is also an airflow mess, especially for the NUC’s, where one NUC’s exhaust is fed straight into the next ones intake.
Repurpose two of the NUCs as hypervisor hosts, you could then create a cluster which contained HAOS, your Ubuntu installation, your sandbox and your broker. Ease-of-management would be on a totally different planet. As well as getting HA (High availability), snapshots and so on.
Huh? HOW is uptime measured in years poor maintenance?
Yes, the wires are a mess. I haven’t cleaned them up after installing a new POE switch which required rearranging stuff. The NUCs are about a half-inch apart and have been running this way for more than two years. This is what lm-sensors in the top NUC (my Plex Server) is reporting:
Core 0: +43.0°C (high = +86.0°C, crit = +100.0°C)
Core 1: +45.0°C (high = +86.0°C, crit = +100.0°C)
This is well within the safe operating temperature.
Patching and maintenance, my son.
???
I update core and OS shortly after they are released. If I don’t reboot the server, the uptime counter just keeps ticking. So how is a long uptime an indicator of poor maintenance?
How about firmware and BIOS? Uptime resets at a reboot.
What about them? As I said, my HAOS NUC receives OS and Core updates when available. A reboot is not necessary.
My Plex server just runs. I see no need to update the BIOS for a working system, or even if Intel has issued BIOS updates for computers this old.
My Ubuntu computer is for experiments and testing programs as well as other Linux functions. A reboot of this computer is not at all unusual. Sometimes I really screw it up and I just restore to the prior evening’s Timeshift backup.
It’s really not something to discuss. If you measure uptime in years you are not maintaining the system.
If everything is up to date- the OS, Core, Integrations, Add-Ons, how does not rebooting the host PC translate to “not maintaining”? You still haven’t said why I would need to reboot the host PC?