The cross section of angles surfaces has a mathematical relationship to the potential for gas vectoring. There's a naval/ artillery whitepaper giving all the formulas. The short answer is that if you want to counteract recoil and not just flip, you have to leave some area for the gas to push against more or less in the way of its origninal vector. The really thin comps do poorly for this reason.
It's rapidly diminishing returns. Basically the first about inch and a half of ports/ vectoring surfaces do work, then the pressure being applied to any remaining surfaces drops off logarithmically. IMO, the comp at the top is made backward, vent most of the working pressure before any of the gills. a thinner walled version has less gill surface area.
I'm from a boat building family, and this stuff works very similar to a rudder. You are essentially blocking a high pressure area with a controlled vent in a new direction. That creates a low pressure, and a fluid nozzle (like a rocket motor). The open ports on the top thrust down, but they don't pull forward significantly, since the perpendicular area on their front edges is very small, and they are wide open. Adding those features after there is no gas left to work with is pretty much cosmetic.