Agree with Geoffrey Beal. However, given the seriousness of the statements, I think that adding quantification will greatly add to the discussion. For instance, are self plagiarism more frequent in predatory journals than in non-predatory ones with comparable impact factors and equal access (or suscription price).
Variable: Frequency of plagiarism/Nr. publications in a given time.
Comparable groups: Same impact factor, same open-access policy, same charge per publication, same number of journals (if it is an editorial group, licke Omics).
This would greatly illustrate this discussion.