Is a kinematic rigidbody preferred over manually transforming gameObject?

If all I need to accomplish is some basic object transformations (e.g. move to another position, change scale) and I don’t need any physics interaction, is there any reason to attach a kinematic rigidbody to the object, or is it less overhead to just perform the transforms manually?

I suppose the second part of the question is, what exactly is the point of a kinematic rigidbody? If rigidbodies are used for physics and the kinematic setting disables physics, why have it at all?

Manually editing object transforms is definitely preferred when the object does not need physics interaction; it will indeed have less overhead.

Setting rigidbodies to kinematic does not disable physics; it essentially “freezes” the object so that it cannot be affected by forces, joints, or collision response (collisions will be detected, but the collision response force will do nothing). This way, its physical behavior can be completely determined by script or animations.