I don't see how comparing the search terms of "XP problem" and "Vista problem" can be an accurate gauge of how buggy an OS is. For all we know, those people might not know how to use/find a certain feature or perhaps a certain program worked in a way they didn't expect.
Also, I doubt that people who have a decent knowledge of how to use a computer would use such a broad search term and that leads me to believe that most "problems" are the result of user error.
There are numerous reasons (including the ones above) as to why those graphs you displayed aren't an accurate representation of an OS's performance.
I'm surprised that you would go along with what is mostly hearsay evidence when you so clearly opposed it when it was pointed at the state of BL after the release of V9. Perhaps we should make a graph based on the number of complaints from users before and after the release of V9?
Also developers (90 percent of them) would rather develop for xp than vista
Yes, because it is the mainstream established choice of operating system for the time being. I am sure those same 90% of developers (wondering where that figure came from) preferred the previous mainstream operating system too before it was made obsolete.
As an example...judging by the graph below...clearly illegal downloading has decreased over the past 4 years!