On another note, does the program keep a history for each metric? In some instances I imagine it would be ok for a metric to go over a threshold but it would not be ok for it to regularly go over a threshold. One that comes to mind is cpu temp. It would be ok for it to peak once in a while for a short time but if it peaks regularly or for long periods then that becomes a problem. And are these thresholds adjustable? Again, for cpu temp 90 degrees might be ok for some machines but is fatal for others. In other words I'd want to know if it was going over 70 degrees.
Coming from a monitoring background, I can give you a bunch of metrics people will ask for.