I imagine the next chapter in this story will be someone scraping the web for public google doc links, reconstructing the keystroke timeline, and training a NN with the timeline in order to break captchas or other keystroke-pattern-based identification systems.