I wonder if I've been under a misconception about low grafting. I thought for cherry trees, and I don't know if this is general or applies particularly to cherries, it was advantageous to graft them below the surface so that they would grow their own roots. Not good? I have recently found some ornamental cherry trees ('Kanzan' and 'Shiro-fugen') with rootstock sucker growth of both the rootstock and the scion. Presumably, sucker growth is not desirable, but it does indicate that the scion roots are not fully inhibiting the rootstock.
From what I've read concerning citrus, the rootstock imparts qualities to the scion such as disease resistance, vigor, and dwarfing. To have the scion add its own roots would work against that.