Hi Robert. I have been trying to incorperate fruit trees into the forests out here in hawaii and it has been a learning experience. In a nut shell there are very few plants that will work in a few situations. As stated, light gaps are an option however there are things to consider and you will not likely establish your trees very easily.
Part of the idea with growing a "food forest" is that you are in control of your resources. This is very important. You start off with open land, ammend the soil if needed, and plant about 90% NTFs and 10% fruit trees. As the system ages, you constantly prune back the NFTs to mulch and feed your desired long term canopy which eventually outcompetes the NFTs and grows from 10% to 90% biomass as the NFTs shrink from 90% to 10%. You mimicked the way nature builds a forest and tweak it to make your own.
Starting a food forest around existing trees will be more difficult because existing trees can send in roots that can outcompete your desired trees roots. Therefore, you are not as much in control of your resources and that is a big challenge.
What I found works in situations like your where you dont want to clear all trees is to figure out what late successional trees can succeed in the present situation. Late successional trees tend to:
1) have dense wood, 2) grow slow, 3) be shade tolerant to some degree, and 4) produce few but large seeds. This is opposed to Early successional trees which tend to:
1) have soft wood, 2) grow quickly, 3) be shade intolerant, and 4) produce many small seeds.
So in Hawaii I found that a few things worked. Avocado successfully grew up through and overtopped common guava when growing in deep soil in valley bottoms. But as a heavy feeder, avocado (and practically all trees with large fruits) did much poorer on steep slopes where soil was less developed. Mango grew up through most forests eventually becoming the dominant canopy, even if it took 10 years or more to reach the canopy. Avocado, mango, and breadfruit all seemed to succeed under large albizia overstory and in theory, one could kill the albizia standing and be left with just the dense fruit tree planting. A few others also worked but the list was generally small. Most trees did not succeed in most situations and were outcompeted by existing tree's root systems, even when fertilized and in light gaps.
Bottom line, to be successful, you either need to clear enough land to be in control of your resources, or you find the few species which can fill the successional nitch and naturally succeed what vegetation is already there.
Thats what I've found.