Okay, I recorded videos and analyzed the data. Here it is:
Method: Videos were recorded of seven matches in a row that included Wu Kong (they can be uploaded for proof, if necessary). Antidotes were brought along to cure any additional blinding effects that would affect the data. The number of hits and misses were recorded when Wu Kong was activated. Of course, data was not collected when Wu was inactive or an ally was dead and Wu couldn’t affect that hero. After the data was collected, it was analyzed to include means, standard error, and significance level. The following table displays these findings.
Analysis: In seven samples, Wu Kong hit 121 times and missed 64 times for a percent missed of 34.6%. This is slightly higher than his card indicates it should be (32%), BUT it falls within the standard error of 3.215 and is statistically significant at p-value of .0121. This reveals that there’s a 98.8% chance that his miss percent falls somewhere between 31.39% and 37.82%.
Conclusion: The evidence points to Wu Kong’s miss percent being exactly what’s written on the card, 32%. The sample may appear small, but the analysis indicated that the sample size was large enough to acquire a statistically significant result well below the requisite p<.05 level. This means that no further analysis is needed because the sample was large enough to draw this conclusion. Additional data would likely make the standard error smaller and it would close in around 32%, giving us an even more statistically significant result.
TLDR; There’s nothing wrong with Wu Kong.