Im not entirely sure what the wet Perimeter would be but your guess seems accurate. I have done a couple Radiator simulations and i havent had to define this yet. I had documented some of this process here but if you have pressure drop VS velocity data then your golden. If you cant get this from the manufacturer (best option) then you can simulate a small Fin/tube section (lets say 50mm X 50mm) and use that pressure loss data as a guide for your porous media darcy and forchheimer coefficients.
Dan