in zainasheff&white's YEAST book they explain pretty well, and show a simple experiment to back up the idea, that as you decrease your pitching rate, you decrease your 'yield factor' (cell count increase per degree plato consumed). practically, that means that as you pitch your pack into larger and larger starter wort volumes, at some point you reach a cutoff where the yield starts to decrease. you still get more yeast out from a larger starter than a smaller one, as there is more sugar and more yeast growth, just that in a huge starter, for each gram of DME you are getting fewer cells in return. in their experiment they pitched 1 pack/vial into different sized starters and measured the growth rates, the optimal results were in the 1-2L range, but from 0.8 to 8 liters were in the same ballpark, and still perfectly acceptable for the homebrewer. of course these rates are calculated for you in yeastcalc.com, as the others have pointed out, play with the numbers and starter sizes on the website and see what works for you! and these experiments were done with a fresh 100 billion cells, whereas shipped packs of course have significant viability dropoff; so if you're really pitching 10 billion cells into 7 L, your yield factor may be really on the downward slope of the curve, and you will really benefit from a smaller step-up. thanks for the excuse to go back and re-read that section of the book.