Don't forget p=0.06 was for one of the pre-specified endpoints (which excluded deaths by undetermined cause).
The primary endpoint missed with p=0.11.
So if the study was designed to accept a 6% p value, it would have still failed.
According to Dr. Ray, the study was underpowered because it was designed to detect a reduction of 30% of events. It could not detect with statistical certainty that there was an effect, if the effect was smaller than this.
And in statistics and medicine, there is no such thing as a "near miss".