Nearly every scientist I know is doing "a bunch of incremental improvements based on a 4-6 month release cycle".

They just call the results of the releases "publications".