I'm a little confused here..... Hal wrote "Excessive dihedral limits maneuverability. "

With a rudder and elevator plane the primary roll response comes from the interaction of the dihedral responding to the yawing that is induced by the rudder. So if you want a snappy rolling model then you want lots of dihedral. If you go back to model designs of the late 50's and early 60's when Class1 for rudder only and throttle and Class 2 for rudder, elevator and throttle the aerobatic designs for these models used closer to 10 or 12 degrees of dihedral per panel and at least one model biplane design used closer to 15 degrees. Sure it looked like a pigeon with broken wings but this was how they built them to work with the class rules. Yes the model will want to return to level flight quicker after the rudder is neutralized with the dihedral but if you don't have the dihedral then it's not going to roll very fast in the first place.

At one point in my modeling life I had a left over 6 foot span Malibu glider. I added an Enya 09 to the nose and took it out to re-aquaint myself with radio control flying after a 6 or 7 year layoff. That glider turned out to be VERY aerobatic thanks to the gross amounts of poly-hedral and a large rudder. Full rudder would yaw the model visibly from it's flight line and then the polyhedral would make that puppy roll like it was on a string. And I don't mean lazy barrel rolls. These were surprisingly axial at a roll rate of about one roll in 1 1/2 seconds or so.

So to my way of thinking you'll want to aim for closer to the 10 degree mark rather than the 5 degrees. And I suspect using a flat center panel with tip dihedral at a higher 15 or 20 degree angle would work even better because you'd concentrate the dihedral effect out at the tips where the leverage would make the roll coupling force more effective.

And incidentally. You can fly inverted with a rudder elevator model very successfully provided you're airfoil isn't too high lift a type. Just use the rudder control like it's an aileron control. Sure the rudder moves the opposite way with the model inverted but then the roll couple works backwards too so the net result is that you have a model that flies inverted just fine but it needs a bit more attention because the dihedral obviously wants to try and roll it back upright so you need to catch it with a little bit of rudder to correct it. I flew a fair amount of inverted flight during a slope soaring session with a poly hedral Top Flite Metric a bunch of years ago also. At first I didn't think it would do it but it wasn't any worse than balancing a broom handle on your finger tip. I ended up figure 8'ing back and forth across the slope inverted the whole time and STILL going up in the strong lift dispite the thermal glider airfoil. Great fun....