If your commentary on welding is referring to the fillets, I believe they are there to prevent a stress concentration due to the sudden change in geometry. I further suspect that the beams are non-prismatic because it is harder to model that way. If what you want to do is prove the capability of the 3D printing process, it is quicker to copy a known good design. Once they get their legs they will likely start re-thinking the basic shapes. Hopefully by then, calculation methods will have caught up enough so that you don't waste a lot of time trying to define all your little non-round, non-straight, non-uniform thickness jiggery-pokery.
My concern is quality control and (more specifically) material anisotropy. 3D printed parts are essentially all weld. I don't want to have to resort to (expensive) welding NDE over the entire part.