There are anecdotes and articles where one kind of tree fruit scion has been grafted to the stock of another kind, perhaps with an interstock, and properties like flavour and shape have been affected (I have not linked to any, as I don't have enough rep yet).
In one case, a pear grafted to an apple has resulted in an apple shaped pear with a flavour in between. And in another case, peach stock grafted to an almond interstock, grafted to a peach scion, resulted consistently in oval shaped peaches - and in the accompanying trees where the almond interstock failed normal round fruit as expected from the scion cultivar. And one last study documented how reliably leaf colour of the stock combined with the leaf colour of the scion, to appear in the seedlings of the scion.
Then there is what is described as the old French technique of grafting "in and in". This is where a scion is grafted to an existing tree, and as the scion fruit exhibits desirable properties of the stock tree, a further scion is taken from the growth of that scion, and grafted to the stock tree. The process is repeated, with all scions of different stages available for comparison on the tree, in theory.
What I would like to know, is whether there is significantly more benefit to grafting a scion to a young new rootstock, as compared to grafting to a grown stock tree, in terms of the likelihood of more evident graft hybridisation effects?