I have a question about the following code: Turning quantum nodes into Torch Layers — PennyLane documentation

The input size of the quantum circuit is [0, 2*pi] for each wire, correct? And the output size of the quantum circuit is [-1,1] for each wire, or? Some use general rotations with three parameters. Is there an advantage over rotations only in the x direction, for example, in the variational quantum circuit?

Thanks!