Your goal is to achieve an approximate doubling of the healthy, viable yeast cell count. So the important thing is the ratio of yeast cell count to wort volume. Too little wort and you have just fed the existing yeast, but they've eaten all of the sugar and nutrients before they've all had a chance to reproduce. Too much wort, and you've got beer.
If you plot the numbers that you see on NB's pdf, you'll see that what you get is a curve, which appears to mimic the numbers from Jamil's yeast book. They're a little different, but they're close enough to make me think it's just differences in experimental methods. (edit: turns out that the experiment from the Yeast book was done with no additional oxygen and no agitation, so it may be that both are correct)
If you have 2 containers that will hold .5L, you could make a liter of wort and split the vial somewhat evenly between the 2 smaller containers. That's what I do when I need to make larger starters. For really large starters, make a small beer (under 1.050), and then pitch the yeast cake from that into your big beer.