That's the whole problem with their "experiment". They took a wildly over-valued item where no one would bid half the reserve, and then drew all kinds of unwarranted conclusions from it. PLUS tons of people knew of the experiment and could have altered the result, knowing they would never be buying anything.
But the kind of shill bidding addressed here (bidders raising prices by placing bids not knowing what other bidders are bidding) is a real problem, but it is dwarfed by the most common bidding problem, which is the auction house raising the bids of absentee bidders when they DO know exactly how much they have bid.
So while this is interesting, it fails to points out that most auction houses are less trustworthy than used car dealers, and that extreme care needs to be taken before participating in any auction.
Bruce