There seem to be two separate topics running in parallel. The initial idea I assumed to be a one of surface texture and acoustic consequences. The second appears to be concerned with mass:stiffness. A third touched on is of ideal gluing surfaces, but this seems to have rightfully been settled with universal consensus.
Before either can be reasonably discussed though, I believe it important to be on the same page regarding scope or range of surface texture, as it seems likely that "sanded" may mean quite different things to different people. "80 grit sanded" can mean light moderate pressure from a hand block or random orbital, which even if you started with a fresh sheet may mean .002-.004" (approximated) scratches by the time you're done. If by "80 grit sanded" you mean however straight off a wide belt sander running at full load removal, this could well leave .015-.020" tracks in your wood.
I never thought to assume anyone could be referring to using wood straight off the TimeSaver. If this is the "sanded" we are comparing against scraped or planed, then I could see argument for potentially meaningful effect. When I think of sanded surfaces though, I think of maybe 180 hand sanding, which in my opinion is around a threshold where further refinement will likely yield little more than academic or principled changes. As with all refinements there is a point of diminishing return, which at some point crosses a threshold of below affecting any discernible change in final properties once put in to application.
Quote:
Wondering if I sand the inside and braces with 1600 grit and get it lovely and smooth wether this improves sound or should it be rougher to excite the soundboard more..
To this initial question, I'd have to say no. Plane or scrape, or polish it up to 12,000 micro mesh if you like. Compared to a reasonable 120-180 hand sanding however, I would say that crediting this with enhancement of sound or excitement of the soundboard is likely a combination of romance and misguided snippets of physics. I would always be open to well controlled studies which may challenge my position, but remain highly skeptical that refinement beyond a "reasonably finished" surface could be demonstrated to deliver any noticeable changes to acoustic or structural qualities in the final product.