I think you're missing the point. It wouldn't make much sense to keep the groups isocaloric but replace the calories from sugar in one group with say, protein

I think you're missing the point. It wouldn't make much sense to keep the groups isocaloric but replace the calories from sugar in one group with say, protein in another. You wouldn't be measuring the same thing any more. The entire point of the study is comparing monosaccharides to polysaccharides and observing changes in 'health' (excuse the vague terminology). Sugar is substituted with a longer chain carbohydrate in well planned research.