This is interesting for sure. I wonder if the different build processes that react uses vs angular would skew this in one direction or another. I think npmcharts is just monitoring downloads.
Can you suggest how the build processes would skew it? For all 3 projects I believe the recommended build method involves a cli tool that installs each package once.
even if the built process is skewing the absolute number of downloads it wouldn't change the rate of change would it? So we can say that React is growing faster than Angular, even if we can't be sure which is bigger.