You're forgetting we have the entire dataset included with the report. The seven holes that Goldie describes are actually separated by up to 2400 ft on the X-coordinate and about 335 ft on the Y-coordinate. These seven holes are not close. As Geosim states in their May 18, 2000 report:
Since the mineralized structures involved do not have any demonstrable continuity between drill intersections and the high grades are not confined to a single small area, it must be concluded that cutting is warranted in this case.
Geosim goes on to say:
To determine a capping grade using the decile method, the highest value of the top percentile containing less than 10% of the metal is selected. In the case of the pyritic zone the cap grade is 17 g/t or 0.49 oz/t.
This is one of the main differences between Geosim and PG. Geosim arrived at a valid high grade assay cap through detailed geostatistical analysis, whereas PG simply pulled his 10oz-5oz-2oz cap out of thin air. If you were to apply the earlier cap of 0.49 oz/t to the dataset used for the August 13 estimate, most of the gold resource disappears.