I'm not sure what Palmer is considering here. Yeast cells initiate carbohydrate storage metabolism, glycogen and trehalose synthesis, due to nutrient limitation, e.g. falling glucose, ammonia and zinc levels. I'm not aware of any evidence that low O2 conditions are a driver here. More a coincidence than an association, I'd suggest. That's not to say what he is suggesting isn't going to work. A good starter can be made without mixing on a stir plate. It just takes more time. A better starter is produced in less time with continuous stirring. The continuous mixing promotes an homogeneous media/environment, where all viable cells have access to nutrients. This is true for up to 72 hours and longer, depending on pitching rate of viable cells. Without mixing, the yeast slurry becomes heterogeneous with some regions having limited nutrient supplies, which promotes stress on otherwise healthy cells. How long to stir? It's important to allow sufficient time for the yeast cells to reach 'quiescence', the resting phase. In this phase, apart from maximum growth being pretty much reached, the cells are packed with storage carbohydrates and they have thickened (less permeable) cell walls making them more tolerant of shock. When pitched into brewing wort, even when the starter is acclimatised to the brewing wort temperature, there is going to be a level of shock response, due to the high sugar concentration of the fresh wort. I recommend stirring until the yeast express clear flocculation behaviour, clumping together during mixing. Then continue stirring for another 24 hours. Then pitch, or store cold for up to a few days before pitching. The stored carbohydrates fuel cell activity during the lag phase, after pitching. Note the stir-plate speed only needs to be set to just keep the yeast in suspension. Quite a low speed is all that's required.