I don't think this is a wholly fair characterization of the paper. The study design is decent, and they seem to be forthright with presenting what they actually found. (For instance, their 4th and 5th studies actually contradict each other w.r.t. the impact of feminism, under a naive interpretation, and they just show us what they found.)

However, it is true that the authors come in with a biased viewpoint, and that's fine if you stick to the data because a very large point of the scientific method is to buffer ourselves against biases of authors. It doesn't matter who they are or what they say; just look at the data and decide for yourself! But when evaluating the summary and qualitative statements, then one has to be aware of bias.

The biggest problem comes when the study authors make statements consistent with their bias but not fully supported by the data (just not in drastic contradiction to the data--often the statement isn't really addressed directly by the data but is flavor text providing one way that the big picture would make sense), and then someone else takes that and "simplifies" by pushing it in a way consistent with their bias and also with the statement (no longer attached to data), and so forth. It doesn't take very many steps of bias-infused telephone-game before "science proves that my bias is totally right".