Well, to return to the original question-"Should I try to have them grow like a peach tree with 3 or 4 main branches? Or do I need to do something different?"- I'd say do something different. Open center pruning was a response to brown rot on peaches in the south- it let sunlight and air circulation reduce the humidity that ruined the crop. The only reason I can think of to do it with apples pears and plums is that you don't like ladders. Modified leader is really multiple leader- you let two or three scaffolds turn upward and compete with the central leader. Eventually, you top them all off. There's more pruning of crossing branches than with with central leader, but you'll get more production out of apricots and plums with the additional branching, and pears just have a natural tendency to grow this way. Bart, you said "some (especially some apples) have 4 or 5 branches radiating out of the trunk on a few different levels so there may be nearly 20 branches spaced in 18" of vertical trunk." I'd take most of those off. Even if you later cut out the leader, you'll want at least 6" between each branch. And you'll eventually cut off everything below 3 or 4'.

To me the point of a central leader approach is that it is easier to assure a strong structure by choosing scaffold branches that are not much more than a third the diameter at the point of attachment to the trunk. Commercial growers use it because it is a more efficient system of harvesting light and it increases the per acre yield- but not to the extent that it should matter to home growers. I usually start out all fruit trees as a central leader and for trees I want as open center I cut out the leader after two or three years. The quickest way to get apples is to train a tree to a central leader and do no pruning except removing branches more than a third the diameter of the trunk until the tree bares fruit. After that you can train the tree anyway you want by taking out the center or leaving it and removing branches to create a more open tree.

These are all very interesting. The Cornell information explains a lot. The french site appears to break the pruning methods by rootstock/tree size, but I like the looks of the M106 rootstock type. Still, the others have some interesting features. The lattice look on dwarves is very attractive, but must be terrible work to maintain. I will probably go with pruning towards a vase shape. Other than some fireblight, disease really isn't a problem, although we like to feed birds in the winter and have a lot of english sparrows, house finches, and gold finches. The house finches and sparrows seem to really enjoy digging into the apples. I haven't been able to convince any of the cats to take up residence in the trees, though. Maybe I can hang rubber mice in the trees? Ha! As a side note, these birds seem to enjoy eating the buds off my blue plum trees in the winter. The plum they like to roost in is nearly stripped clean in the spring and has only scattered blossoms, mostly on lower branches. I always have some sunburn on the apples on the south side of the trees but it isn't really a problem and doesn't appear to do any more damage than to hurt the eye-appeal of the apple. I was concerned that maybe there was some production or ease of care advantage with central leader systems that I just couldn't figure out. When trying to maintain a smaller size to a tree that wants to grow, do you periodically take out one of the scaffolds or cut it back severely to spur newer growth lower down (then train one of the new shoots that come out as a replacement scaffold)? Otherwise, it seems that eventually you would end up pruning branches back to roughly the same place every year and have a tree that looks like it is on steroids--big, thick limbs, no smaller fruiting branches?

The easiest approach for quickest fruit with many varieties is just to thin out oversized branches and do nothing else until tree begins to bear fruit. For varieties that send very little secondary branches from scaffolds it is traditional to head the scaffolds back by 30% annually as needed to increase such branches- best done at least a week after green tip. I'm thinking about putting small notches in front of dormant buds where I want secondary branches instead. Commercial growers often use Promelin- a synthetic plant hormone that subdues apical dominance and encourages more branching in a trees interior. Where the second tier begins depends on the rootstock and inherent vigor of variety- less vigorous varieties should be closer and may even need more branches (4 on first tier) per tier. With trees on 7 to 111, 3 feet is about where I start my second tier so I wouldn't head the leader below that. Ideally the tree will produce enough branches that you don't have to head the tree until it reaches max height. Try notching to put those branches where you want them instead of heading. Scaffolds on second tier ideally disect (come in the middle) of territory of first tier scaffolds in terms of directional positioning. That means if a 4 tier structure points NES and west the tier above would be something like NE, SE, NW, SW. Some varieties, like Fuji, that are very brushy may benefit from early thinning of surplus scaffolds instead of waiting until tree begins to bear fruit. These varieties excessively shade themselves which can delay flower formation.