It seems your evaluation is mostly focused on voting fraud. But another harmful effect might be unhelpful voting. Voting is essential in gauging the quality of questions and answers, and it is important in giving feedback to users asking and answering questions.

New users might not be as experienced in what the objective of Stack Overflow is, what makes good questions and what makes good answers. In fact, it would really surprise me if new users would be equally good at this as experienced users, because I sure didn't know how to recognize high-quality questions when I started.

It is not sufficient to only asses if fraud doesn't increase. It is essential to assess that voting remains equally correlated with post quality.

Metrics that could be used as a surrogate for true quality metrics are voting by experienced users, close votes and deletion.

It seems your evaluation is mostly focused on voting fraud. But another harmful effect might be unhelpful voting. Voting is essential in gauging the quality of questions and answers, and it is important in giving feedback to users asking and answering questions.

New users might not be as experienced in what the objective of Stack Overflow is, what makes good questions and what makes good answers. In fact, it would really surprise me if new users would be equally good at this as experienced users, because I sure didn't know how to recognize high-quality questions when I started.

It is not sufficient to only assess if fraud doesn't increase. It is essential to assess that voting remains equally correlated with post quality.

Metrics that could be used as a surrogate for true quality metrics are voting by experienced users, close votes and deletion.