Here's a reference for the 40% bit: http://www.telegraph.co.uk/science/scie ... t-men.html
I'll try and find the original paper as well..  http://brill.publisher.ingentaconnect.c ... 0/art00002
I can't find any evidence in the paper to suggest that male attraction goes down when the percentage of bare skin increases beyond 40%. In the first place it only makes the distinction <20%, 21-39% and >40% (noting also that there was no exposure over 50%, nor less than 10%). If anything, signs simply point to "less is more".