2015 study finds that the glycemic index varies widely depending on the individual: https://www.washingtonpost.com/news/to-your-health/wp/2015/1...
Also, the GI would still be relevant even if it varies, because what you really want to know is the relative GI, and the ordering should at least be preserved across individuals, right? Like if watermelons have a much higher GI than nuts in the table, you're probably not going to find someone whose personal GI is higher for nuts than watermelon?