I've been digging into the topic since "progress studies" got more popular, but it seems like no one has done much in that direction (which is also what my init

I've been digging into the topic since "progress studies" got more popular, but it seems like no one has done much in that direction (which is also what my initial hunches suggested should be tried). There's a lot of grunt work to do if you want to label the data to match our current technological reality, but in the abstract, it still seems to me that building such models could yield interesting insights. You can add all sorts of bells and whistles to make it more interesting (play with diffusion speed, the inherent difficulty of the discoveries required, etc.).