A cross-section image from high-resolution transmission electron microscope shows the interface of epitaxial strontium tin oxide (Sr3SnO) and cubic yttria-stabilized zirconia (c-YSZ) atop a silicon substrate.
(Source: North Carolina State University)
You are right. All sorts of EMI interference will be a factor in systems that substitute spin-state for charge as their binary encoding, but first they have got to get the spintronic transistors actually working!
Shielding would seem to be an issue if you are running switching currents nearby. Won't you get the natural magnetic effects interfeering with the electrons? Might be a silly question- just let me know...