I don't think you need a set of scales to participate in this experiment. If you have access to your own eggs, join us in the experiment.
Jak, thanks for sharing your stats. The telling data will be if you take that same hen who hatched 8/10 roos, and compare her egg shapes. Same with the hen who hatched 4/0 pullets. If you can consistently identify a specific hen's eggs, and then compare the shapes, and leg band the chicks, you'll have some telling data to proove or disproove the theory. Obviously we couldn't expect to tell for sure if an egg is male or female based on shape. I can tell you that in the past 3 years, 2 random hatches resulted in 40% pullets. 3 hatches that were set with "gender selected shape" eggs resulted CONSISTENTLY in a 60% pullet hatch. Call it hokus-pokus, random chance, old-wive's-tale. Call it what you want. I can tell you that I will continue to eat my pointed eggs, and set the rounded ones. If I end up with an extra 20% layers, I'll be happy. Will this always produce more pullets? Hard telling. That's why a growing number of us are playing around with it. However, if I can increase my odds by 20% over time, IMO that's a worth while gender improvement to strive for!