For what it's worth, we worked pretty hard to make this as statistically solid as we could with the dataset that we had. The subject we're posting on is "How 4,000 men differ from 4,000 women in RescueTime". While the data might lead to some interesting questions, I don't think anyone would assume that this is necessarily a reflection of a broader population (any more than Mint's data was).
I think your "black people" example is interesting. If we were comparing designers to developers, would you be so incensed? Yours is literally the first actually angry response all day. Is it possible that you're overreacting?
In terms of business effect, signups today have been abnormally high and cancellations have been slightly to the low side.
It was meant to be provocative, but so far this hasn't ended up being a mine field. Again, sorry that we've upset you.
Here are the following red flag passages, for me, at least:
"From what I can tell, the 23rd chromosome has a pretty amazing impact on the way people use computers." -Drawing a conclusion from data you admit has large sampling issues.
"Women spend more time socializing and shopping" - You did not mention any comparative "frivolous" activities that men might take part in more than women; say, gaming.
"Evidently, there’s a reason they are called “man” hours." - Using a limited data set to large conclusions about female working habits in all contexts
Generally, the whole thing was both hyperbolic and quite inflammatory.
I know YOU don't believe that women are 77% as productive as men and that explains the pay gap, I get your tongue in cheek tone.
The problem is that there are lots of people who use these half cocked "pieces of data" as part of larger arguments and it is exacerbated by a statement like "a random sample of 8,000 people" when clearly, this is not a random sample. It should say "this is not a scientific sample", because it isn't.
This is how we got birthers, "logical racists", and conspiracy theorists.
Perhaps I am more sensitive to it, I took a class on gender studies, I hung out with some feminists, my wife makes a lot of money (more than I for a while, and she wasn't on a computer nearly as much as I).
I think I just get mad at misapplied statistics, the lack of distinction between correlation and causation, and people saying ridiculous things to get attention (of which this does all 3). Suffice to say fivethirtyeight.com is one of my "read every day" sites.
This is not the boy scouts. You don't get a research merit badge just for showing up and crunching some numbers.
Is that professional? Is that good for your company? If I were an investor I'd be PISSED. If I hadn't canned my account to your site about 6 weeks after I made it (cool tool by the way, just missed too much of my work since I don't spend all day on my computer), I'd be canceling it today. And I sure as hell won't ever be trying it again. And I'll specifically be recommending against it.
At my startup we had a discussion about using data from our users to draw attention the way you are trying to here (and Mint.com has very successfully done in the past as just one example). The problem is you have to recognize where the mine fields are.
Replace every single place you used the word "women" and replace it with "black people" and see if it is offensive to you yet.