According to the Food Defect Levels Handbook (FDA.gov):
Refers to the results of the Howard mold count method which is reported as the percentage of positive microscopic fields that have been scored as either positive or negative based on the presence or absence of a minimum amount of mold hyphae.
Many web pages describe the method as "standardized", but I haven't found a clear statement of the standard online. An article in an industry newsletter goes into more detail about evaluation of tomatoes, which are the fruit for which the method was first introduced:
The measuring cell into which the tomato paste is filled has a grid-shaped measuring field divided into 50 equal fields. Each grid field containing microscopically recognizable mold is evaluated. The Howard mold count is given as the percentage of positive fields in relation to their total number.
If blackberries, raspberries, etc. are evaluated in a similar way, then the presence of as few as 30 mold hyphae in a sample might be judged unacceptable, if they happened to be 30 different fields. On the other hand, the FDA document refers to a "minimum amount" and not "any amount", so the threshold for scoring a field as positive may be higher.
In any case, "average mold count is 60%" refers to small quantities of mold contamination seen under a microscope, and not to a product that is 60% mold.