What is the typical setup for switching between third and first person, in terms of model culling?

I have tried a setup in which I had a skeleton identical to my character but only a mesh for the arms, being enabled when switching to first person. The idea was, to avoid the camera clipping issues, and to be ale to work some camera magic to see a point of view from the "eyes" but also have your hands holding your weapon in your screenspace, even though the skeleton was holding the weapon more in front of its chest than its eyes.

This worked, but I figured it would be easier and more efficient to take my original mesh, and assign a separate material to the arms and legs, so the rest could be made transparent on command (i.e., when in first person).

This way is friendlier to the animator blend tree setup, but I have even less control over individual aspects of the viewport and manipulating the screenspace. Additionally, where the transparency stops, might as well be a clipping place, as there are no solid or interior faces for the arms or legs.

So, I wonder, To continue working on this problem, I have to know, what is the typical setup for a camera / model workflow in which, when switching between third and first person views, the model (arms / legs / sometimes torso) is still visible, while avoiding clipping through things like the chest?