Sadly, the authors ignored the body of work that defines the NACA airfoils. Sigh. Months in the lab could have been saved by a day in the library.
Genetic algorithm? Get real. The thermodynamics and fluid mechanics of air were well enough understood to produce a simple airfoil specification in the 1930's.
The authors compare a 15% thick airfoil with no camber to a 16.65% thick airfoil with camber, but they could easily have generated the NACA specification of their new airfoil if they'd just bothered to understand the NACA airfoil specification. If they are going to build a fluid dynamics model in the computer, then they can use any NACA specification they want: "NACA2417" fits their arbitrary new shape rather well and it is as easy to plot as the NACA0015. Once they have the new NACA spec, they can use the NACA fluid models to make predictions of airfoil performance (lift curve slope, angle of zero lift, etc.)
The chord isn't going to be pulled out of the air. If it's too thin, the blade is so flexible that it ripples. If it's too wide, the blades are very heavy and the supporting structure blocks the wind. You'll have to work out the range that is acceptable, and stay away from extremes that are likely to fail.