The rules were formulated in such a way that it was Randi himself—not the panel of independent scientists—who in the end would decide if the evidence was strong enough to reward the prize.
The edicts, reportedly formulated by a very clever lawyer, entailed that
the parties had to ‘agree’ to the conditions, thus giving Randi the ability to ‘disagree’ if a serious contender should appear—not just nutcases or people wanting their 15 minutes in the limelight.
The candidates also had to waive their right to sue. So if a candidate should feel that he/she had demonstrated the agreed upon phenomenon but Randi refused to approve the result, there would be no possibility to try the case in a court of law.
Moreover, Randi appears to have sabotaged applications he found threatening.
The Dutch researcher Professor Dick Bierman, a physicist and a psychologist was once in negotiations with Randi to set up a long-term experiment involving multiple participants. A trial of this type would have been regarded as having much more scientific weight than a single phenomenon produced by a single person. But—when things started to become serious Randi suddenly disappeared without leaving a trace (stopped answering Bierman’s emails, etc.)—just as magicians have a tradition of doing!