I wish I knew where this misconception about taproots came from. Trees (99% of them, excluding oaks and pines) do not have what most think of as a tap root that will stretch many feet under the soil reaching groundwater. Yes, seedlings punch down a 12" root which will seem dramatic in proportion to the aboveground growth, but it does not continue. This is a survival mechanism for seedlings to get a deeper root system than most weeds so they can compete and survive in their first few years.
If you have even looked at a tree blown over in a storm you can see there isn't normally a taproot, just roots spread out laterally. Most trees roots are in the upper 2-3' of soil (including the 'taproot', extending out well past the drip line. Here's some info from the University of Florida extension service:
http://edis.ifas.ufl.edu/mg089.
I'm not sure what your climate is like in England (I hesitate to say my knowledge is the stereotype that it is always raining, maybe that's just London?), but in Pennsylvania where I live orchards are normally not irrigated in any fashion after establishment, so using seedling rootstocks for better access to soil moisture isn't going to be any better than a purchased rootstock unless you are in a dryland, and even then I'm not sure. I guess that the advantage to using seedling rootstocks (since it will be a semi wild area) is the dwarfing factor shouldn't be too strong, but will exist as grafting naturally dwarfs to some degree.
If money isn't too much an issue, I'd look into getting rootstocks that impart disease resistance over starting your own from seed. Starting from seed will also put you 5-7 years from fruit instead of 2-5 years from a purchased rootstock.
On a semi-related note, the reason it's not recommended to start fruit trees (especially apples) from seed is that you can plant 100 seeds and maybe get 1 variety that is good, 5 that are mediocre at best, and 94 that are basically crabapples. Modern fruit breeding gears itself towards mechanization and shippable fruit, but the basic fact that there are many, many duds before a good variety remains true.