Is there any difference between AddForceAtPosition and AddForce + AddTorque in terms of performance?
I would think that they really don’t add too much in the terms of performance. I am sure someone closer to the inside could tell you if when you call those functions if it just keeps an array of events to be decoded in the physics step, or it adds that to the current number that is there. I would suppose if you are calling those functions 1000 times per frame, it could bog things down.
I made some tests. AddForceAtPosition looks like slightly slower than AddForce + AddTorque.