Wow! You've done a neat job of building up a new rudder from scratch.
The old rudders have a nasty habit of doing exactly what happened to yours - the sideways forces on the rudders are pretty extreme. I've had at least two or three of them go on me. Hobie "fixed" the problem in the late 70's by introducing Lexan rudders. They didn't break, but they flexed like crazy and you couldn't get a good, smooth finish on them.
Then came EPO rudders - foam core epoxy fiberglass. Far and away the best rudders ever made. Stiff, light and virtually unbreakable. They have a hard finish that would take a high polish. Even though they've been out of production for at least 15 years, they are still highly sought after.
The current racing rudders made by Hobie are nearly as good. My only complaint is that the gel coat finish is soft and scratches / dings easily.
I've seen several hand-built rudders over the years - from beautiful, all wood creations to high-tech all carbon fiber. Each have had their problems.
In looking at your photos, the biggest concern I would have about the way you've constructed your blank is the hard transition between the wood and the foam. That's going to create a shear line, which if it falls on the bias of the fiberglass cloth, will be a significant weak point. (The bias of the cloth is diagonal to the weave - it is the weakest part of fiberglass.)
Going back to my college statics engineering, the rudder can be simply modeled as a cantilever beam with a uniform load:
Nomenclature:
R = Reaction force
V = Shear force
w = Distributed loading force
M = Moment
l = beam length
x = location
E = Modulus of elasticity
I = Moment if Inertia
y = deflection
Relationships:
--------------
R1 = w * l
M1 = -(w * l^2)/2
V = w * (l - x)
M = -(w / 2) * (l - x)^2
y = (w * x^2) / (24 * E * I) * (4 * l * x - x^2 - 6 * l^2)
ymax = -(w * l^4) / (8 * E * I) at x = l
As you can see from the graphs, the moment and shear are at their maximum at the base (left side) of the cantilever. That's why your old rudder broke off just below the casting.
In reality, it's more complicated. The load is not uniform; the beam is not uniform; and the three dimentional effect of twisting is not taken into account.
Good luck on your rudder! Post photos when you are done - let us know how much it weighs. Give us a performance review - who knows? You may have come up with the latest and greatest new design!
Matt Bounds
(one-time naval architect)