They aren't making a noticeable change to the system at all. This method challenges roughly 350 patents a year while roughly 300,000 patents are granted per yea

They aren't making a noticeable change to the system at all. This method challenges roughly 350 patents a year while roughly 300,000 patents are granted per year. That's in the area of 0.1% of all patents. I think these numbers are very consistent with the predominant use case being the challenging of bad patents and there are very few examples of people trying to overturn good patents.