>The DX equivalent is rather 16mm at f5.6. In both cases - FX >and DX - a 36" print would be diffraction limited, a >24" print would not be.
I would think that softness due to diffraction, as is DOF, would be dependent upon amount of enlargement and viewing distance. The often quoted f/16 for FX and f/11 for DX is dependent on an 8x12 viewed at about 18".
Assuming the 24" and 36" measurements are the long side maintaining the 3:2 aspect ratio and keep the viewing distance relatively normal for the enlargement size, then the f/16 and f/11 would still define the relative sharpness. But if you view the 36" enlargement from the normal viewing distance of the 24", or the 24" from the 12" viewing distance than the affect of diffraction would be evident,as well as a change in appearent DOF.
On the other hand, a diffraction limited system is dependent on the pixel pitch and as such is format independent. For example most telescopes are diffraction limited. If you want more detail for a given focal length you need a bigger opening. On the other hand a good number of lens are aberration limited and that's why they generally perform better stopped down a bit. But at some point stopping down further causes lower contrast so the diffraction spot becomes the limiting factor. And if the spot size becomes three times the pixel pitch, the system becomes diffraction limited in ability to resolve detail. We can correct the softness due to enlargement of the Airy Disk through deconvolution. But resolution lost through diffraction cannot be regained.
But honestly the difference is really small, and due to the limits of human acuity, I can't help but wonder if this all doesnt to some extent or another fall under the old adage "How many angels can fit on the head of a pin".