The cross section is the important constant in RMH design. What we have learned from all the years of trial and error is that a J-tube system with a given diameter that is pretty much equivalent across the burn tunnel, riser, and flue will perform well as long as other dimensions are in alignment with the ratios that are known to work. I'm intentionally citing diameter as the constant, instead of cross sectional area, because another thing we've learned is that as you get farther away from cylindrical sections, the effects of friction on the fluid dynamics are amplified.
This is not to say that you can't push the envelope in one or more directions. But the riser is quite possibly the least tolerant of diameter variance out of all the main parts of the system...the gases in there are at their most expansive point in the process. A riser that is too constricting will impede the flow of combustion gases, choke the draft, and inhibit clean burning. At the other extreme, a riser that is too wide will decrease the velocity and turbulence at the critical stage where the highest temperatures are developed, and this is also likely to limit mixing, efficiency, and clean performance.