I believe the argument for using a stir plate is that it does continuously oxygenate the wort if a vortex is achieved.
The only time that a stirred starter begins to approach the surface area of a Shaken, not Stirred starter is when a vortex is created. The vortex is needed overcome the low amount of surface area that is present when a 1L starter is made in a 2L Erlenmeyer flask as well as to overcome the geometry of a Erlenmeyer flask by creating a vacuum. A stir plate cannot overcome the poor geometry of an Erlenmeyer flask once the culture starts outgassing.
When we use a stir plate to propagate yeast, we are using a process known as stirred suspension cell culturing. Shear stress is a well-known problem with stirred cultures. Shear stress is the stress placed on cells in turbulent flow. The more turbulent the fluid, the greater the amount of shear stress. I am convinced that the foul odors and tastes that one encounters with a stirred starter are the result of stress, not oxidation, as oxidation should not occur in a culture due the cell density and yeast's affinity for O
2.
Shear stress in stirred cultures is a well studied problem. While we think of growing yeast biomass when the term culturing is thrown around, many other types of cells are propagated using the same basic techniques that we use for yeast (the medium may be different). A device known as a bioreactor is used for very large scale cell culturing and dry yeast culturing. There are numerous papers covering this area of research.
Mark's point is that the forces present in a starter that is stirred at such a high speed will cause damage to the yeast cells. It seems it's not so much a biochemical problem but a physical one. Apparently all that slamming around during long periods of stirring is detrimental to the cell walls of the yeast. The shaken not stirred method does employ very vigorous/violent stirring (shaking) but it is for a very short period of time (1 min) such that there is really no physical stress to the yeast. I take his method to heart and really shake the living daylights out of that one gallon jug to the point that the whole 1L starter is completely foam. The maximum degree of air-to-liquid surface area is achieved in that condition.
Yes, the key to success with the method is shaking the medium vigorously enough to turn it completely into foam (or at least darn near it) in a vessel with at least a 4:1 vessel volume to medium volume ratio. Gas dissolves into a liquid at the interface between the gas and the liquid. The Shaken not, Stirred method works by creating a huge amount of surface area. A gas-liquid foam has a very high specific surface area. It is basically pockets of gas encased in thin layers of liquid. The beauty of the technique is that the O
2 that dissolves while the medium is still foam is available immediately to the culture.
With the above said, shaking can be eliminated via direct O
2 injection. However, the technique is no longer low tech and low cost (or as my British friends would say, no longer as cheap and cheerful).