Benchmark

[CarsonBurke]

At least one benchmark should be created to be able to compare the performance implications of changes, and test the limitations of this library.

I'd argue that an example should be created where:

• many lights are created and destroyed periodically, say every 3 seconds a few hundred new lights are spawned
• some lights are moving
• some lights flicker
• some lights have extremely low falloff rate, spreading very long distances
• many occluders should be present, blocking lots of light

Text UI should be present that displays the current FPS, 1% lows, and a moving and/or stable average.

I am comfortable doing this if necessary/desired.