[initialsBB]

From the study:

Combined, the four major labels sold an average of 491,000 tracks per week in France after March 2009. Our findings indicate that in the absence of HADOPI, if France followed the same trend as the control group,20 sales would have averaged only 401,000 units per week. Thus, our results suggest that the HADOPI law (and the education and media attention surrounding it) increased iTunes single sales by 90,000 units per week on average. If we assume an average song price of €1 per song, this equates to an increase of €4.7 million ($6.3 million) in annual iTunes track revenues.

also

Accounting for the fact that our independent variable is in log terms, these estimates indicate
that French track sales units rose about 25.5% in the control group after March 1, 2009 but by 48% in
France, indicating that French iTunes track sales were 22.5% higher on average than they would have
been in the absence of HADOPI.16 Similarly, album sales units rose by 42% in the control group but 67%
in France, indicating that HADOPI increased iTunes album sales an average 25% per week in France.
Standard errors are clustered at the country level, and these difference-in-difference results are
statistically significant at the 95% confidence level.17

I don't understand statistics well enough to know what's going on in their graph. But what you seem to be arguing is that the study's own graph contradicts their findings. I think that's highly unlikely and I suspect that you're really just misreading the graph.