The front of the embedded horn section would be in full contact with the back of the beveled hingeline stock...the hingeline stock would remain uncut. The embedded part would carry through to the bottom sheeting. The resulting surface area of the embedded assembly would be greater than a normal hardpoint, plus the added benefit of contacting the hingeline stock and BOTH sheeting surfaces...not just the bottom.
Click on pictures for full size:
The sheeting is blue, the foam core white, the hingeline stock green. The purple would be lite ply with matching lightening holes. I've seen people screw or 'pin' these types of assemblies together (and I've done it also on big horns) but I can't imagine the shear strength of good ol' thick CA being exceeded without ripping the entire assembly out of the surface anyway. If you installed them with polyurethane glue they'd be PLENTY strong yet much lighter than other options.
Obviously you lose adjustability, but who needs it? If you used a 1.25" horn you'd have 1:1 mechanical advantage...if you needed to decrease throw / increase mechanical advantage, go in on the servo arm holes. You shouldn't ever increase throw / decrease mechanical advantage by decreasing the control horn length / increasing the servo horn length (resulting in a mechanical advantage of LESS than 1:1) anyway, right?
The solid model shows the assembly being flush with the outer upper skin. For appearance sake you could remove some material from the assembly and backfill the slot with balsa, then sand smooth to the upper surface for a truly invisible installation.
Let me know what you think. If I had a bazillion of these things made they would be less expensive than anything else available.