A fair chunk of your initial datapoints are challenging that assumption.
One could say the same for the $20k minimum: there are certainly people who make $5k or $12k also, but he's grouped them together into essentially "make very little money", just like he's grouped $200k+ into essentially "make a lot of money".
At the moment, people making $200k or more are 15% only.
Stick a >$300k on there, and I wouldn't be surprised if a big slice of that group shifts right over to it.
Incidentally, in response to _delirium, we really need to stop thinking of $200k as "a lot of money". That's an achievable wage for a good developer. Describing it in terms normally reserved for fantasy rather than reality is not particularly helpful.
Sure, you can make more than that, but in a general discussion about freelancing, it's not all that interesting to break down fine gradations of the upper-middle-class, which can be usefully bundled into the category of "people who are very successful at consulting/freelancing". Heck, I'd probably put the cutoff at $150k+ if I were designing a poll, since above that threshold you are no longer trying to make it as a full-time consultant who can support a good lifestyle.
I do think it could be interesting to break out the truly wealthy, adding a separate category for $1m+. But there isn't really any interesting difference between $200k and $300k.
Bill G would be seen as "one of us" Larry Ellison much less's so.