Russian trolls on Twitter had little influence on 2016 voters
washingtonpost.com
washingtonpost.com
Seems like all the political influence action was on Facebook. Twitter was just more "public" than the "walled garden" of FB.
For example: at the time, almost half of Republicans said they don't think Obama was born in the US, and another quarter said they didn't know. Only 1 in 4 Republicans could agree on that basic fact, long after the man had released the long form birth certificate they had demanded. (I haven't seen any polls more recent than that, but given that it was already 4 years since Obama had left office I doubt it has changed much.)
The misinformation is clearly out there. The only thing we learn from having it spread by paid Russian trolls is to have it identified as misinformation; they have no interest in spreading actual facts. (Or if they say things that are true, to put them into a context that is misleading.)
The trolls mostly just repeat the kinds of things that are repeated on right-leaning media. They're ineffective, but only because they're only telling people what they already believe. It doesn't mean that people don't believe false information, only that they've already heard it.