Did you look at the data? Because I don't think you did. I lay it out here[0]. In short, what we can see is that in 2021 the 55+ age group has higher support than the 35-54 group did in 2012. This is not people aging into the new bins and bringing their support with them. If that were what was happening, then the 2021 55+ group could not have greater support than the 2012 35-54 group. There is missing growth from your model. We're literally talking about 50% of the growth being unexplainable by your model.

[0]https://news.ycombinator.com/item?id=39087434