I used to have the same thought about starting plants in the largest pot to avoid repotting, but have found that that is the easiest way for the plant to develop root rot. When the soil to root ratio is high, the soil in container will stay wet for too long since there is not enough roots to use up the water.
Planting in the earth is different because the earth acts like a giant wick and draws the excess water elsewhere.
So I would go with repotting. It is sometimes even appropriate to repot a size down if root pruning is implemented.
-Herman
I think it is safe to say that the conventional wisdom is to go with gradual up-potting.
I agree with Herman but I would offer a different explanation. He brings up the issue of what I know as the "perched water table" (PWT). The PWT is the saturated growth medium that still holds water in the bottom of the pot after watering. It is held there by the adhesive properties of the water. The growing medium in this portion of the pot is anaerobic and the plants are vulnerable to root rot if this perched water table is too high.
The height of the perched water table is dependent particle size of the growing medium. Taller containers will have a smaller
volume PWT as a percentage of their
total volume. But the actual height of the table is dependent on the growing media particle size. Perlite/pine bark/Turface/coconut husk chips (CHC)/turkey grit have large particle size and minimal PWT. Compost has small particle size and as many of you know, after a time gives you that stinky anaerobic sludge you sometimes see in the bottom of a container.
The particle size in most growing medium tends to break down over time. Peat, and pine bark, common elements in mixes, will break down fairly quickly and the particle size will be reduced. (Slower to break down would be perlite and CHC, and Turface and grit may not not break down much at all.) Smaller particles wash to the bottom, and over time the PWT in a particular pot increases because of these smaller particles.
Herman's results, in my opinion, may have to do with the fact that by potting to the largest pot, you keep that plant in a given potter longer, giving those particles a longer opportunity to break down, and raising the PWT of the pot. By potting up, you keep consistently larger particle size in the bottom of the pot and thereby a lower PWT.
I will also offer one other observation of the benefit of gradual up-potting. I think that the root density with up-potting is increased and more even throughout the container. It is my own observation that when placed into a too-large container, roots will sometimes crowd the perimeter, leaving a loose, crumbly portion in the center. It is hard to root prune these plants because if you remove the outer 1-2" of root ball you have removed 90% of the roots!