This is really interesting. I wonder how difficult it would be to have a set of "training texts" where the reader/user is presented with different parameters to

This is really interesting. I wonder how difficult it would be to have a set of "training texts" where the reader/user is presented with different parameters to observe optimum points for speed vs retention. Excited to read the study when it's published.