Parameter inverted in insaff branch

[andreww]

Report from Alan: "I just tried your MS_effective_medium insaff option, and I may have found a bug. I think your znh input parameter, which you say is equal to zn/zt is actually zt/zn (the reciprocal)." - presumably this is the unmatched insaff branch.