Sure, but there are so many Steam users every month, that even if only 1% or 0.1% of people get the survey, that's still a ton of people (over 40M concurrent, estimated 130M+ users every month)

Also, have the Linux users increased by over 50% month-to-month? The jump from ~3% at the start of the year to over 5% is huge.

It's possible there was something wrong with the data before and the 3% wasn't accurate or there's something wrong now and the 5% is not correct. That's all I'm saying.