I have done the capacitor fringing problem in a somewhat more direct
way than a spreadsheet relaxation procedure. But what I have done is
get the field from two sheets of charge. I beleive that the capacitor
with plates of finite thickness will have some of the charge migrating
around the edge of the plate. Bur the spreadsheet ought to handle that.
I usually try to use Mathcad but Mathcad will do a relaxation
problem only if it has a square boundary. I does sound like things
could get pretty complicated wspecially if you want field lines
as I do.