I played around a bit and came to this model that should serve as a starting point for you.
There were recurring mesh failures that I traced to the perfect alignment between the lower horizontal tail and the mesh on the fuselage. Shifting that tail up to Z=0 made it so the mesh did not align and we got a mesh that ran successfully.
I set it up to run with the wings as Thin and the fuselage as Thick. Unless you have a strong reason to do otherwise, this is my recommendation.
Don't start by running 21 angles of attack. Debug everything with just one. Once that is working well, then add some more as needed. I promise you do not need to run 21 aoa.
You had it set to 64 wake points -- this is probably what was causing the biggest slowdown. I backed it off to 8 and down to 4 wake iterations. Start with small numbers on parameters like these and increase if you have a compelling reason -- don't start high.
Likewise, I decreased some of the resolution of your mesh (and clustering). It is still probably more than I would work with, but I was trying things until I could get solutions working. Once I had things reasonable, I didn't keep playing around.
Rob