He should've studied psychology if he wanted to understand human sexuality, there's a whole literature on it. I guess that would mean having to understand statistics and how research is actually conducted. Let me debunk some of his 'psychological' claims.
TL

R Ben doesn't know what he's talking about when it comes to the psychological nature of pedophilia.
Therapy for any anxiety disorder, including PTSD, does include exposure therapy. A patient needs to willingly and carefully expose themselves to their fear. As such they grow in confidence, they don't lose the fear itself. But that's for anxiety disorders not sexual attraction. There is no hard 'cure' for sexual attractions of any kind, just suppression which is not the same brain mechanism for anxiety and overcoming it. Anxiety is avoidant behavior while sex is approach behaviour. From the victims point of view, lolicon isn't the fucking 'trigger' of their PTSD, its fucking abuse so exposure to lolicon wouldn't work anyway.
I agree that age play and pedophilia are not necessarily the same thing BUT are definitely connected. Peadophilia is attraction to the prepubescent form. Clearly an adult, no matter how hard they try won't have the same body structure as a child so in that sense they're distinct. Being aroused by an adult form even disguised as a child is not the same as arousal from a child. But clearly, trying to attain the form of a child for sexual gratification is in service of prepubescent attraction. He's right pedos aren't attracted to the childish performance but the prepubescence itself but then says it is not the sheer form of loli that attracts pedos. This is an out and out contradiction in two sentences. If you like lolicon you're a pedo. Pure and simple.
Where does he get the data for only 1% who watch cp commit offences against minors. First of all, participants lie all the time. Second, a well known phenomena in psychological research is the responses bias, where participants tell researchers what they think researchers want to hear. Where do they find pedophiles to question in the first place, it's hard enough to find participants for any study. An outlier in statistics is a data point within a data set that seems wrong, an error or just a participant who is very distinct from the other participants in a given study. A reaction time of less than 100ms for example. However, researchers aren't allowed to remove that data point from the statistical analysis unless there is very good reason for it, a reaction time of 0 or a reaction time that is unreasonably long, otherwise that data point stays in regardless. Plus, he misses the point, that in order to produce cp those making cp are committing an offence against a minor.... So it really doesn't matter how many pedos watch cp and never commit an offence. For every video, an offence was committed. Just because two things correlate doesn't mean they are related. In fact correlations are mostly used in research to justify further investigation with the data and not used as conclusive evidence. Of course he will have a 'correlation' he has what you call a sampling bias. All the people he talks to believe the same thing so the participant pool he's drawing from are skewed one way.
Finally, the pictures he present as 'an exploration of child abuse' are not drawn in an horrific way meant to 'shock and horrify', they're drawn to attract.