>>He has the power to stack ODAC committees with an inordinate number of biostatistician voters.<<
I wouldn't say that he stacks the ODAC with statisticians. There are typically 10 to 15 people on an advisory panel, with only one or two statisticians, but he certainly has the ability to chill an audience in a general meeting (e.g., one on surrogate endpoints) by calling out the person's name and company as they approach the microphone to ask a question.