Everything reverts to the mean and your point is valid - we shouldn’t expect one form of writing to have abnormally high returns to effort ratio for too long, unless there is a general shift in the value of writing by the population at large. Which given current trends towards video and higher stimulation, I would assume is less likely?
Unless one could argue there’s a huge untapped latent market of bookworms. I think there is some market to a degree as people get burned out on other forms of media, but I’m not sure if it’s large enough for substacks model to work for a significant period in the future