Isn't the resolution on these governed by the magnetic field strength (like the nuclear spectrometers in chemistry?) That's why MRI machines are so big; the magnets are insane (and (edit)[involve] a bizarre material: liquid helium!)
If, however, you use some significantly more sophisticated signal processing, you can side step this requirement.
Worth noting: A lot of the focus on advanced signal processing has been focused on scan speed, and main field strength and homogeneity is only one factor in resolution. You need high quality gradients too, the finer the resolution the more precise everything has to be set up or you are just smearing the signal around in space.