> Perhaps I was wrong to expect mention of explicit discsussion of spectral graph theory?

Agreed... if you're going to write down the Laplacian at least demonstrate the null space <-> connected component relationship.

Edit: looks like maybe an undergraduate paper?