Practically speaking if there was an efficient muffler for c/L no one would use it simply because of size and weight. Aluminum is not a real good dampening material for noise, the choice for most is light steel and sufficient volume to slow everything down before it exits the stinger.
Bob Davis at Davis Model Products supplied some of the most efficient German made mufflers to the r/c community and had very little luck with them because of the same objections as the C/L crowd. That said Europeans fly large gas burning models with no more noise then the family sedan because they have to comply with strict noise restrictions. most have very well designed steel mufflers with efficient baffles in them. Regretably efficiency is probably not pretty on a stunt model.
What would I judge as efficient. When you can fly in a populated area and not be chased because your a noise nuisance.
I think that pure exhaust noise is somewhat overshadowed by prop noise and this muddies the waters in regards to what is heard as an 'efficient' muffler.
I have seen straight through all aluminium mufflers that work on the absorption principle wrapped with glass matting that are very good at noise suppression - the only down side is having to repack them from time to time . If you want a really quiet and simple system then extend a perforated pipe from your exhaust, wrap it in fine glass cloth and slide a covering tube over that - no need to seal off the ends of the outside tube if it is a good tight fit.
Light, easy to make and has very little back pressure, but as I said you will have to repack it every now and then due the weight gain from oil absorption.
The Davis muffler, was that the reverse flow three piece cast job that emulated the BMW motorcycle design? Totally over engineered in my opinion if it is what I am thinking of!
No matter how quiet your model ends up, if you fly control line in continuous cyclic circles there will always be the repeated Doppler effect that will drive people nuts. The human mind simply hates the same noise rising and falling over and over again (this effect is the basis of most alarm systems) no matter how low the Db reading is.