My beef is that you changed the denominator to try and raise the ratio of people who could logically be said to know of her. It's not about improving the model as this didn't do that. This just reduced the sample size down to Instagram users instead of the US population. It isn't a remotely fair extrapolation to say that because some percentage of Instagram users have heard of her then that ratio must be the same for the US population. That's not going to be the case because the entire point of social media is getting followers and likes and searching people out...but whatever. The bottom line is in the scheme of things, hardly anyone knows who she is. Regardless of whether it's 1 in 10,000 or 5,000 or 2,000, she is not well known in any context of "celebrity" status in the US, and certainly not in the rest of the world.
The issue I have with her post is that she took something someone said to her privately and used it for the sole purpose of creating drama to create publicity for herself. I don't have too much of an opinion on the comedian analogy because almost all of those are made up or modified so much from whatever hypothetical conversations they had...they are paid entertainers. Here we are assuming she didn't make this up (which I think is certainly debatable and the context isn't known). The fact that the person remains anonymous doesn't mean it's not public. Think of how many times you see "sources said" or "an official who wished to remain anonymous said..." on a media platform. Those are very much public domains even though the source (like the guy here) stays anonymous. Who knows how many people she told where she named him...or maybe it's all just made up?