Having conducted a reasonable amount of academic and scientific research, this metric is more likely to be mischaracterizing research than revealing any issues.

Having conducted a reasonable amount of academic and scientific research, this metric is more likely to be mischaracterizing research than revealing any issues. This doesn't even establish a causal-link between self-citation and poor research quality, it just assumes it. Most researchers continue to do new research on the same concept after a publication, and they will of course site their earlier work when continuing. Additionally, post-graduate researchers often have their names placed on the research of grad students they are in charge of, even though they often have minimal involvement in the research or conclusions drawn. You might be able tell something from the ratio of other authors from all citations to the number of self-citations, but only if you could eliminate self citations that were not either inclusion by proxy or cases where they are merely continuing research on the same topic with new methodologies. There are already methods for identifying bad research, none of which can be achieved through the use of non-human-assisted data analysis of the authors list of research. The only way to be sure is critical review and 3rd party verification of results with repeated experiments.