What we see here is that there are some very dubious twitter accounts with some interactions. But are these meaningful interactions? Did any of the content reach and influence the supposed target audience? Are the communities shown in the Gephi graph real communities? These communities and their interactions could just as well be fake. When I was working on exactly this, I noticed that fake communities (i.e. sets of fake accounts all retweeting each other) are quite common.
Just look at the specific accounts mentioned (Pascal66616113 and 1Happeningnow). They have only 2 and 51 followers, respectively, and right now it's 4 and 72.
I think it's very important to detect malicious political accounts on social media and be aware of their impact. This article mentions two examples of such accounts, but doesn't prove anything about impact. The conclusion that France is losing an information war here doesn't follow.