The military also have teams of cognitive scientists A/B testing different HUD apps. I remember seeing one paper where they were A/B testing different presentations of incoming enemy fighter craft. This stuff is tested by people who have very deep knowledge about perception and attention, and aren't just an app slapped together in spare time.

I remember from that testing that there were three ways the threat was categorised by distance: safe, lethal, very lethal - the middle one was very specifically not "kinda in danger" :)