A very fair question.
I is not very common to have a lone N-pole or E-pole on your mesh, especially if your mesh is closed.
But if you happen to have a mesh with a single E-pole or N-pole, the usual unpolling techniques will move them around or even eliminate them.
In the next gif animation, I move the only E-pole present by the same unpolling technique. If you keep moving the E pole, it will reach the edge of the edge and vanish.
which results in this:
With each unpole you introduce more edge loops. You may want to remove these extra edge loops afterward.
To remove a N-pole you should follow the chart below:
Just cut along these red lines and join triangles.
To continue where I left off, I was working on the nose area. There are some standard topologies widely used for the nose. It all depends if you want a anatomical correct solution, or a topology that deforms well etc.
The red loop should be always there IMO because it will make the nostril look like it sprouted from the face. The nose should be somewhat in its own topological region, separated from the upper lips in any case. The green loop is what I see a lot in the ‘old days’. It ends at the eye. I find faces with this setup look very CG-like.
The purple loop will form a C-loop. It will form a nice face with well defined cheek bones. This loop travels around the eyebrows.
The blue loop will lead to a X-topology which is just awesome. A lot of very professional looking models has this X-topology. It always seem that the wire frame is light and clean.
But in this case I won’t be aiming for any loop in particular because the nose area is so messed up… so I’ll take it from there and I see from there where it will take me.
To begin with, the mouth loop flows right into the nose
Let’s get rid of the mess on the side of the nostril first:
step one: move two key e-poles around
Some action in the animation above may seem new. Well, another way to move a E-pole around is by selecting those 2 rows of edges and then collapse them (Mesh menu–>edges–>collapse). This method decrease the mesh density instead of increasing it when using the former method. In this case I think less mesh density is better because you have a better overview of the base topology.
As you can see, the loops at the tip of the nose are forming nicely, but the loop that goes around the chin-mouth-nose is a little messed up. This is easy to fix: