I'd suggest that the "stronger" is one of those "it depends" sort of things. Balsa is weak in one direction but it's actually darn strong along the grain. Meanwhile plastic sections of the same weight will tend to be brittle.
Mind you much could be done by using capped I and T sections for the printing. And since you appreciate that the whole model can't be done I like the idea of sections that go together and then have carbon rods, tubes or strip stock slipped into place and bonded to take the actual loads. That sounds like a rather workable option.
The neat thing is that with the right option the sections could be make already "covered" and with slick panel lines and other details molded in place. Joints between sections could fall at the joint lines for the full size surface panels and lip and tongue joints would ensure a good fit and maximum gluing area.
The sub structure that these sections are assembled over could be a mixture of carbon rod, tube or strip for longer structural members and plywood plates or built up box sections for connectors and specific load fixings.
All in all it's a slick sounding option. But I feel that until they can produce fiber filled layups to enhance strength that the optimum use will be as "covers" with carbon and plywood internals that join and carry the loads. That would allow the printed portions to be as light as possible consistent with the need to be strong enough for normal ground handling and basic flight loads.
The slick thing is that pushrod guide tubes and hinges can all be printed right in place. Flaps and ailerons would come off the printer platen already hinges with just the need to connect pushrods to the molded in place control horns. Complex Fowler flap tracking on smaller and lighter loaded models could similarly be printed right in place since they would not require the sort of reinforcement that metal pins and the like would provide.
Of course what time we save at the building board would be more than taken up by the poor guy doing all this CAD work....