Good evening aLTeReGo!
Thank you for your reply. I do understand your workaround solution and I am going to set it up, to get a reaction on the (general) system health overview - where it is possible (since not all servers are ready for this monitor).
I agree, it is working that way around. But it is a pity, because the values for disk space or even physical/virtual memory usage do indeed affect the node health state.