Users can see the granular information, such as how many "likes" included a comment from the user -- only 6% on Mashable.com, I found (knowing that the comment functionality on the Like button doesn't always work). I also was able to learn that the "likes" that included comments were more likely to get clicked on Facebook, which makes sense because it adds a personal voice from the user. What is interesting is that Insights also breaks out the shares that occur using the plugins and "organic" shares, which essentially is how many shares took places by users simply copy-and-pasting links to share or using the old Share button.
This is likely to bring some great insight to designers who implement Facebook's "Like" buttons to see how its placement affects the click-through rates of the button.
Demographics for Interactions
Insights now breaks down interactions based on anonymized demographics, which may enable site owners to target specific audiences.
Publishers will be able to learn more about the audience that interacts with its Facebook plugins on their site. For example, site owners can see what age group, gender, language or country is most active in using these plugins on their sites. However, publishers should note that the data for demographics shows numbers per interaction, not per user. This means the numbers shouldn't be interpreted as showing how many users are interacting with the content on the site but rather how many interactions are taking places from a specific demographic.