"Scope: ON" and "Scope: OFF" tests measure the scoping performance in M3G
These tests cannot be evaluated individually, only in comparison to each other.
In both scenes there are 100 meshes and 2 lights.
In "Scope: OFF" test all meshes and lights are in the same scope, so every object is lit by 2 lights.
In "Scope: ON" test one third of the meshes are lit by only the red light, one third lit by the blue light and one third lit by both lights. On average one mesh is lit by 1.33 light. If scoping is not a bottleneck "Scope: ON" should be at least as fast as "Scope: OFF"
The result in both cases equals to the number of rendered frames.