Yes but how do those convictions happen? If it comes from an interweb honeytrap, then it will give somewhat unbiased data. But I suspect a lot of people caught for possession are people at the fringes of a abuse case where the cops couldn't prove this individual actually abused someone, but could get them on the lesser crime.
And how many charged with child molestation will have the police raid their homes and check their computers? How many will end up only being charged with child porn to spare the actual child the trauma of going through a case and how many times is there not enough evidence to convict for child molestation but there is plenty to convict for child porn?
Admittedly, I don't have the study on me, so I can't confirm this. But I imagine the study would screen out anyone like that, that seems like a pretty obvious thing to do.