Interesting...so you're saying that pitching a starter earlier (before fermentation is complete) would reduce the oxygen requirement (i.e. oxygenation of wort)? So for someone who does not oxygenate, they would be better off pitching earlier rather than later?
It's always better to pitch at the end of the deceleration phase than it is after fermentation is complete. Nothing is gained by allowing a starter to ferment out. No net increase in yeast biomass occurs during the stationary phase, ergosterol and unsaturated fatty acid reserves are depleted, and the cells are subjected to increasing amounts of fermentation byproducts, including ethanol. Allowing a starter ferment out also results in the cells being in the yeast equivalent of hibernation.
At the end of fermentation, yeast cells go into survival mode where their cell walls thicken and they store carbohydrate as glycogen. In effect, the cells are preparing for hard times. It takes longer to exit this state than its does when the yeast cells are still in active growth mode; hence, lag times in addition to oxygen demands are also increased.
There are only two scenarios that I come to mind where it is okay not to aerate a batch of wort. The first scenario is when pitching a quantity of yeast that is large enough that the cells do not have to undergo much in the way of multiplication. The second scenario is when pitching dry yeast.
Dry yeast is propagated using a continuous process that is very different than that that is used to propagate liquid yeast. Dry is yeast is propagated aerobically in a device known as a bioreactor (a.k.a. chemostat) where the glucose level is kept below the Crabtree threshold via continuous injection of medium (usually molasses with supplemental nutrients) and removal of yeast. Aerobic propagation is a significantly more efficient process because no ethanol is produced. Aerobic propagation also leads to yeast cells that have fully charged ergosterol and unsaturated fatty acid reserves due to continuous injection of oxygen. Liquid yeast cultures are usually propagated in batches where the glucose level significantly exceeds the Crabtree threshold (the glucose levels found in beer production exceed the Crabtree threshold by a large margin).