The real source behind this statement was that Ok Cupid once published a blog post with some analysis of their data. They found that the average man on the platform was rated as bellow average attractiveness by woman while the average woman was ranked as average attractiveness by men. There were some more related stats in the same post.
This of course exploded and they deleted the post.
This one?[0] It's funny it's always cited with the reversed finding: while men were more fair in rating attractiveness, they then only target the top most attractive females, completely oblivious of where they stand. OTOH women were more harsh in rating but they are then able to shift their expectation to realistic ones. By these data dating only works if you are in the top X% of attractiveness if you are a woman, but it works better for the average man.
Nice shade. But if someone is able to meet people in-person and get dates from people in that top X%, then the conclusion that it's the online component of online dating that's at fault doesn't seem unwarranted.
I find that most people that say this only attempt to match with people in the top X% of attractiveness.