that simulation is with one half of the main wing loaded, basically what happens when you do a full wingout turn just before you crash 
. It doesn’t model the downforce from the tail, so in reality there would also be some stress on the aft part of the fuse although it would be almost all direct downforce with only a small torque component. It’s important that the rear fuse is stiff, so while it can be made thinner (low torque/side loading) it should still be tall.
a lot of the size of the mast foot/fuse is there to accommodate threaded inserts and to keep the fuse stiff. The F4 solution is to make the whole foot into the threaded insert (conceptually similar to the old Nolimtz), it’s a good way to reduce profile but the engineering and production of the metal/carbon interface has to be really good. Epoxy doesn’t stick to metal very well and the different thermal expansion rates of metal and carbon/epoxy means the bond will most likely fail at some point so there needs to be a mechanical bond of some sort to keep things in place.
I have ended up with something that’s sort of a cross between your idea and the F4/Nolimiz concept. It’s a short metal fuse that’s permanently bonded to the mast with the mast slotted into the fuse and bonded and pinned in place. Bonding means that there’s no slop and tolerances aren’t a problem. There are no metal inserts in the mast and the fuse can be replaced if necessary due to stripped threads/corrosion or whatever.
