This is MSI (Meaningless Statistical Information) often used by politicians to convince the less informed. They just omit the part that makes their statement stupid.
1 - It could be that there are more users now than when they first run the program, so in total numbers WX wins, but in percentage it fails.
2 - It could be that their customer satisfaction program wasn't aggressively pushed before, so less people answered it before, giving WX the nod now.
3 - It could be how they asked the question. For example, "Are you happy WX is free?"
4 - It could be that the people that really hates WX skipped the survey because they were frantically trying to return to W7 and had no time to waste with stupid surveys.
5 - Or... it could be that people that waste time with satisfaction surveys just enjoy wasting time with WX.