They probably shouldn't put the average amount paid so far on the website. The idea is to get people to pay what they think the games are worth, not what other people think the games are worth.
http://images1.bingocardcreator.com/blog-images/hn/deep-disc...
Anyhow, the idea was to simplify the workflow and simultaneously abuse the power of defaults and user psychology. I also intentionally ditched some parts of the offer which I don't think add business value, such as fine-grained control on the split.
(I'd A/B test including "65% of people pay this much" on the #1 step, probably with a subtle yellow highlighting.)
Edit: As others point out, there are already some other anchoring points on the page, so having an average lower than those may be hurting them. It might have been wiser to make the "Top contributors" tab the default (or only) tab in their realtime stats display.
I also notice in the promo video it has the average right next to the entry form. On my form it defaults to 29.95 and says "For instance: 10, 20.50, or 100.00." as help text. It seems like they played around with the form a bit.
This is the type of thing that could benefit from a nice A/B test. Might be hard to run on a very public one-time promo like this though.