As a first step I'd use the normal dynamic clearance +3 mm on the radius.
I don't know how to calculate a viscosity,I'd build 3 or 5 with different viscosities and fine tune by ear. You could construct a dynamic model to optimise this, but in my experience it would be indicative at best.
The inertia ring size should be related to the modal mass of the mode(s) you are trying to suppress. Practically speaking there is usually an upper limit that can be packaged, so design it to that and then hollow it out later. Again, this can be modelled, with some degree of success. There is a trade-off here - adding mass to the damper will reduce the bending frequency of the crank, which may hurt you more than the benefit from absorbing the torsionals.
Cheers
Greg Locock