Use a more accurate FastACos implementation

Assets/ScriptableRenderLoop/ShaderLibrary/Common.hlsl

 
 #define MERGE_NAME(X, Y) X##Y
 
+// Acos in 14 cycles.
-    float res = -0.156583 * x + HALF_PI;
-    res *= sqrt(1.0 - x);
-    return (inX >= 0) ? res : PI - res;
+    float res = (0.0468878 * x + -0.203471) * x + 1.570796; // p(x)
+    res *= sqrt(1.0f - x);
+
+    return (inX >= 0) ? res : PI - res; // Undo range reduction
 }
 
 // Same cost as Acos + 1 FR