Solving Fat using zCloth is difficult to handle especially if the range of motion is more.
a. I agree Technique 1 is bit tedious to handle.
b. Technique 3 as mention was used before pressure, Rest scale, surface tension introduced in tissue. This techique is also good but after the those parameter added in tissue. It eliminates usage of zCloth. Which in turns directly helps in sim time.
c. Technique 2 and 4 are similiar only difference is that in one we are coupling Mus and Fat together.
I use both the technique. for initial testing of muscle and fat parameters i use technique 2. Where i can iterate more and play around with the parameters, material layers , attachment etc.
its bit linear approach solve Mus first than Fat second. Once happy with the output. Use the zbuilder and copy and paste parameter of muscles and Fat in one file. And take the full advantage of the bi-directional pull.
I don't worry about the interpenetration between muscles. But yes self penetration is one u need to watch out.
By the way in Bi-driectional technique self penetraton issue gets minimize. While testing the lion cub sim it really helped where they are playing n wrestling a lot .