Males are a greater risk than females by a 10 to 1 ratio
Policy make sense to me
Let's try to guess this.
Say the probability that an adult is a "kiddie fiddler" is 1 in 100,000. (For an Australian population of 22,000,000 and assuming 80% are adults over 18 years, that means 176 adults in Australia are child abusers). To put that in some perspective, if we flat assume there are 170 pax per flight, in 588 flights, we expect one of those pax to be a "kiddie fiddler" (within all of those 588 flights).
Now that 1 person we pinpoint is 90% likely to be a male (if we take the statistical conclusion from the analysis you linked to). So, by extension, 0.9 pax in 588 flights, which (using whole numbers is easier

) translates to the likelihood of a male "kiddie fiddler" being on board is 1 in 654 fully loaded flights. (For reference, 1 in 654 is about 0.15%). For a female "kiddie fiddler" this translates to 1 in 5882 flights.
There are a number of assumptions which actually make this estimate quite conservative. And yes the stats is not quite perfect - this is just back of envelope, over-a-coffee/over-a-beer talk stuff, but I'm trying to illustrate a point here.
The question is then, if this is the risk of actually having a male "kiddie fiddler" on board who may have the potential to abuse a UM (
if a UM is on board that particular flight), is this risk significant enough to deem it unacceptable, and thus this policy is absolutely necessary? Yes, the odds are higher for men than women, but the overall risk - is
that unacceptable? A risk of 9 in 10 flights having a male "kiddie fiddler" on board is more than likely unacceptable......
Note that you can't have zero risk - the only way to do this is to remove all males on board a flight which has a UM (i.e. the risk is eliminated completely) - this also means if the crew are male, they go too!
Note that the statistical approach above also completely disregards the moral, practical and legal implications.