Have you read a book by Jon Ronson called "The Psychopath Test". It's an easy read, and underscores the difference between an actual sociopath (/ psychopath), and people who have sociopathic traits. It partly involves and discusses research that suggests a lot of people at the top of society have traits similar to sociopaths (or narcissists)... but really aren't.
Sociopathic traits exist on a spectrum, and we're all on that spectrum somewhere. Full-on sociopathy is generally assumed to be a childhood development issue and pretty much fixed by the time of adulthood. However, we also have to consider some sociopathic-like behaviours are learnt behaviours people have adopted later to fit their circumstances and roles, because such behaviours can work. (For instance, one might argue the function of some military training is to induce some sociopathic-like traits). But at core, underneath these learnt behaviours there's a "normal" person who could revert back where a real sociopath could not.
I think there's a fairly gossipy trend to diagnose people from what we see of them in media, but unless maybe they are particularly obvious and grotesque (such as Donald Trump's narcissism) it's very unsafe to make these sorts of assumptions. Ronson's book illustrates the ease with which we see a few traits that mislead us to make someone fit a profile, when deeper analysis would reveal major problems with the diagnosis. For instance, a major and common issue for sociopaths is impulsivity, and there is little evidence Holmes was impulsive (compare that with, for instance, Trump's narcissism-induced, hair-trigger impulsivity). Likewise she doesn't appear to have had particularly unusual problems forming decent personal relationships (friendship or romantic), which seems inconsistent with sociopathy. And, simply, Holmes behaviours could be understood and explained for other less dramatic reasons: ambition, insecurity, even naivety, etc.
