In general there are always multiple AC algorithms available which all do it slightly different.
For S2 data there is, ICOR, sen2cor, MAJA and others, even C2RCC has its own AC over water.
They take into account for example, Water vapour, shadows (mountain cloud), sun/view angle and more.
The DOS algorithm is so to say a poor-mans AC.
So, you actually can’t compare the DOC algo with the real AC algorithms.
Different ACs have been compared and will be compared in a second round.
https://earth.esa.int/web/sppa/meetings-workshops/acix (related paper)
https://earth.esa.int/web/sppa/meetings-workshops/acix-2nd