The incident meter "assumes" that everything in the photograph is 18% grey and reads accordingly.
The reflected meter "converts" everything to 18% grey.
So, it's question of priorities. What are you trying to achieve in a given photograph?
With the fox in the snow, a spot meter reading of the fox will give you a pretty good exposure, without any bracketing, because the fox's coat ( I am assuming an orange/red fox of the type we have here in Ireland AND that the sun is behind you.) is as near as dammit to 18% grey.
On the other hand, the (far easier to take) spot reading of the static snow will give you an exposure about two stops underexposed, since the rendering of textured snow is about two stops above 18% grey. So, you will need to adjust the reading accordingly.
But, an incident reading of the scene will immediately give you a working exposure which, given the assumed dominance of the snow within the scene, you can tweak accordingly. i.e. bracket -1/3, for example.
As to your example of the shadows: well, how important are they to the overall image? In bright mid-day sunlight, with the sun directly behind you, a "normal" ( and I use the word advisedly) exposure for ISO 64 is 1/250 sec. @ f 8/11. Now, let's say you take a spot reading of the shadows in the scene and they are five stops below. Will they hold? Probably not. Are they important enough to overexpose the rest of the scene a little? That's up to you.
You have to prioritize what you are seeing.
I don't think you're missing anything! ;-)
ps. Have you read Ansel Adam's "The Negative"? It's all there! ;-)