Fatigue failure is a function of the magnitude of fluctuation in stresses, in addition to the peak stress loads. Worse would be fully-reversing loads, not the case here as slender spokes will not support compression loads. And, number of stress cycles, which equates to the distance the wheel has rolled. But reduce the stresses, and the fatigue limit gets pushed way to the right on the graph.
I'll accept a lower-spoke wheel if the rim then is thicker and stronger at the spoke holes, I think I have seen this on some rims, so it's not a pure extrusion but is formed or machined to be thinner between the spoke holes.
Otherwise, double eyelet/socket rims are the magic bullet for rim durability, if you don't mind the extra rim weight.
I think on a thread about either an old Lemond or Armstrong race bike, a closeup of the rim showed the spokes exiting, not at the center, but on each side, with a hole halfway between inner and outer diameter, and I immediately thought, that's a great design, because a) the spoke hole area is loaded more in shear than bending normal to the surface, and b) with the rim overall in bending, that halfway point is near the neutral axis in bending, a really good place to attach something. Also, the spokes crossed laterally, from rim right to hub left, and vice-versa, which increases the lateral spoke angle to improve lateral stability and improves truing I think.