Yep. It's all about the testing to see which works. Back when I flew some .15 diesel control line combat the hot prop of the day was the white nylon Tornado 8x6. But the prop didn't really come into its prime until it had been boiled for 1/2 an hour to "temper" the nylon so it wouldn't break on the first crash and then to thump it in a few times so it was bent back at a jaunty angle and THEN it would really let the engine work.

Of course all of this sort if means this thread is really out of context within the "Aerodynamics" forum where it's all about the theory of making this stuff better.

Is there a better prop design waiting to be discovered? Likely there is. Will it be head and shoulders above the present optimum? Not a hope. A few %? Yeah, there's probably a little room. What'll it take to get there? Likely about 50-50 computational study and practical cut and try. The computational part suffers to some extent when dealing with our unique sizes and the aerodynamic issues. The best issue would be to study what we have now and then model the computational stuff until we obtain the same results as in real life. Only then once it has been calibrated in this manner can we alter the blade shapes and airfoils and study the results. Even then if the shapes change radically then you'd need cut and try to check out the computational results.

As always when you're doing this stuff it comes down to GIGO.... Garbage In = Garbage Out