The questioning of Sheldrake's credibility is a bit of hyperbole--he is actually a fairly reputable biologist, who advocates that science is a method, not a position-

He also contends that a lot of what passes for science is merely habit, and his critical point is that scientists tend to think that double blinds are only needed when there are human subjects involved--to cancel out their bias, and are completely oblivious to the idea that experimenter bias has a great effect on the outcome of research.

This is not a new idea, not even a controversial idea--most scientists generally forgo acceptance of what seem to be new discoveries until they have been independently verified many times.

The trick is, they recognize experimenter bias can be a significant in the work of others, but not in their own work.