There is definitely a long-term societal benefit to doing both kinds of work. Often work seen by only a few people (which might gather only a handful of citatio

There is definitely a long-term societal benefit to doing both kinds of work. Often work seen by only a few people (which might gather only a handful of citations, or none at all) inspires a later generation's breakthroughs at the frontier of human understanding. (For example yesterday there was a post about Hermann Grassmann and his pioneering work in linear algebra which got almost no positive feedback at the time but 150 years later is considered foundational for whole fields of study.) On the other hand making very popular material that helps a large number of people to slightly better understand old well-known ideas can still have huge benefit.