What I hadn't heard before was the part about stationary cells being reported to be more stress tolerant.

The morphological changes occur in order to harden the cells against the harsh reality of living in a toxic, low-nutrient medium for an unknown amount of time.

For what it's worth, I believe the Mr. Malty calculator uses the fairly standard rate of .75 million cells per milliliter of wort per degree Plato (and twice that for lagers).

I'm not saying that's the be all and end all of pitching rates or that it isn't worth experimenting with different rates for different styles of beer, but it's not like Jamil just arbitrarily made it up.

Cell count is only half of the equation. Cell health is equally, if not more important. When pitching cells at high krausen, one is pitching very healthy cells with very pliable cell walls that require little to no maintenance before they can go to work; therefore, reducing lag time. The number one bogeyman with normal gravity fermentation is house infection. Shortening the lag phase shortens the time that bacteria have to multiply (as does reducing the number of replication periods), and bacteria multiply three times faster than yeast. These growth rates are what we are up against in a real world brewery.

https://www.homebrewersassociation.org/forum/index.php?topic=19850.msg277460#msg277460"A small amount of bacteria can overtake a much larger amount of yeast because the bacteria cell population increases 8-fold every time the yeast cell population doubles. If we were to normalize the propagation period between yeast and bacteria (bacteria multiplies three times faster than yeast), the growth equations would be:

yeast_cell_count = initial_cell_count * 2

^{n}, where n = elapsed time in minutes since the end of the lag phase / 90

bacteria_cell_count = initial_cell_count * 8

^{n}, where n = elapsed time in minutes since the end of the lag phase / 90

If we run the numbers, it should become crystal clear why one wants to pitch a large, healthy yeast culture while doing everything possible to minimize the opportunity for bacteria to catch a ride into one's yeast crop, starter, or fermentation vessel. It should also become clear why the growth phase is called the exponential phase.

Cell counts at 90 minutes

yeast_cell_count = initial_yeast_cell_count * 2

^{1} = initial_cell_count * 2

bacteria_cell_count = initial_bacteria_cell_count * 8

^{1} = initial_cell_count * 8

Cell counts at 180 minutes

yeast_cell_count = initial_yeast_cell_count * 2

^{2} = initial_cell_count * 4

bacteria_cell_count = initial_bacteria_cell_count * 8

^{2} = initial_cell_count * 64

Cell counts at 270 minutes

yeast_cell_count = initial_yeast_cell_count * 2

^{3} = initial_cell_count * 8

bacteria_cell_count = initial_bacteria_cell_count * 8

^{3} = initial_cell_count * 512

Cell counts at 360 minutes

yeast_cell_count = initial_yeast_cell_count * 2

^{4} = initial_cell_count * 16

bacteria_cell_count = initial_bacteria_cell_count * 8

^{4} = initial_cell_count * 4096

Cell counts at 450 minutes

yeast_cell_count = initial_yeast_cell_count * 2

^{5} = initial_cell_count * 32

bacteria_cell_count = initial_bacteria_cell_count * 8

^{5} = initial_cell_count * 32768

Cell counts at 540 minutes

yeast_cell_count = initial_yeast_cell_count * 2

^{6} = initial_cell_count * 64

bacteria_cell_count = initial_bacteria_cell_count * 8

^{6} = initial_cell_count * 262,144

Cell counts at 630 minutes

yeast_cell_count = initial_yeast_cell_count * 2

^{7} = initial_cell_count * 128

bacteria_cell_count = initial_bacteria_cell_count * 8

^{7} = initial_cell_count * 2,097,152

Cell counts at 720 minutes

yeast_cell_count = initial_yeast_cell_count * 2

^{8} = initial_cell_count * 256

bacteria_cell_count = initial_bacteria_cell_count * 8

^{8} = initial_cell_count * 16,777,216"