There are a couple of threads here where people tested this, have you seen them yet?Recently this has been the case for studies one of these studies is A morphometric method of sexing white layer eggs http://www.scielo.br/scielo.php?script=sci_arttext&pid=S1516-635X2013000300006
I've been thinking about a method of testing this method reliably with white leghorn lines. Crossing a Silkie rooster with White Leghorns, hatch the eggs based on shape and see what is the actual percentage of hatched females(this specific cross makes it easy to sex them the moment they hatch)
I'll find them for you if you haven't.