First, I'll admit the question was poorly worded, but folks still seemed to get the general idea.
Secondly, I get the feeling people are reading far too much into this last poll question. Similar to the comment on the original thread criticizing the poll for having an obvious Windows bias. This is a Windows site, and the polls aren't deep and accurate. They're 1 question posted for a week.
These polls aren't scientific. They're also not trying to paint a comprehensive picture. To get a complete picture of OS usage of our users, I would have to issue a complete survey, with multiple questions carefully worded to not influence answers. I would then have to do some actual analysis of the results beyond simple % values. Sorry, but figuring the margin of error on these things is probably not worth the time.
The polls are here to spark some conversation, and give at least a partial idea of what the community thinks about a topic, or what they use on their home PCs