Nice article. FWIW I'm not sure you are using the short data correctly.....you are initially looking at the change in cumulative short interest and dividing that by number of trading days to come up with a daily average over the period but then you are comparing that to daily short figures. The issue here is that most of the shorting is done on a daily basis and not held overnight, hence large daily figures but not much change in the cumulative short interest at each reporting period. My hunch is that the daily figures are fairly similar and consistent in the months prior to the dates given in the chart you posted. Do you have that data?
(This is definitely not meant to suggest there wasn't manipulative short selling around the previous SA articles, just that it can be very difficult to decipher).
Market Data 
Markets 
AI
