[betterunix]

"These smaller, more digestible, incremental papers are also more useful to other researchers"

Survey papers are useful to other researchers. Minimum publishable units are a waste of research time, and the pressure to pad one's CV and get "results" (which actually means more papers) worsens the situation. It is even worse when a grad student is told that they were "scooped" because their half-finished MPU was published by someone else.

[coliveira]

MPU papers are useful for training purposes. They let students and junior researchers participate on the scientific conversation. It is important to have minor roles to be played in the larger scientific community.

[XorNot]

Personally I hate the entire format of publishing a paper. It's more spinning a story then actually detailing useful information, and the single longest exercise of my PhD program (which I dropped down to a Masters from) was a result of a group publishing many papers on a subject and never actually detailing their method properly (turns out the whole thing is dependent on partial oxygen pressures, so the detail of "lightly placing the lid of the vial on top of it" - mentioned nowhere - is spectacularly important).