If light.target.properties.targetInverse is false, no targetLine and
targetSphere is created, we can't update their values on update()!

Removed Vector* add() to addVectors() and addSelf() to add(). Same with sub, cross, min, max, lerp... Renamed Matrix* and Quaternion multiply() to multiplyMatrices(), multiplyQuaternions() and multiplySelf() to multiply(). See #2860.
Good thing we have a ton of examples to make sure nothing is broken. The unit tests also helped. However, I suspect I've left something unchanged.

Added .add(), .addColors(), .addScalar(), .multiply() and .multiplyScalar() to Color. Removed ArrowHelper from DirectionalLight too.

I think it's a bit redundant.

Also, code clean up here and there.

Based on http://filmicgames.com/archives/75

Add missing space
Fix isIntersectionSphere
Addunit test for isIntersectionSphere

Tets if two spheres intersect

WebGLDeferredRenderer: added API for injecting postprocessing effects into deferred rendering pipeline.

The problem this tries to address is - how to incorporate surface normals into diffuse term area light calculations. The original approach completely ignored surface normals, which resulted in both light shining on backfaces and lit surfaces not showing any shading. This made prettier empty box with very soft look but anything inside the box had completely flat diffuse term.

Not yet fully solved. I tried myriad variants, all methods I could come up with had some problems. This one for example breaks down when the light emitter area intersects object triangle :S.

Also fixed handling of vertex colors for emissive term and tried a bit closer match with shininess strength between area lights and other light types, though this is tricky, seems it would require some non-linear mapping.

Frustum.contains -> intersectsObject, containsSphere->intersectsSphere, removed Frustum.containsAnyPoints, .transform.  Update unit tests accordingly.