The search for "temporary fix" picks up a lot of snippets like the following, which is code that manipulates temporary files (e.g., files in /tmp that are removed after a process exits), not "temporary fixes":
Also, most temporary fixes probably aren't explicitly commented as such. So this search is picking up lots of things that are not temporary fixes while probably missing the vast majority of the real ones. Thus it looks like the methodology for this survey is quite flawed.