The S2 and O3 filters would behave the same as the Ha. Each would focus at a slightly different point but that happens anyway. I believe the pixel scale would have an infinitesimal difference between the three masters.
I was specifically looking at a particular 152mm f/7.9 ED doublet as the basis for building a new scope. Based on some of the published ray trace diagrams for lenses at this size and f/ratio, the RGB focuses are separated by about 1 mm. This means you have a focal variation of 1mm in a 1200mm focal length. I see a lot more focus spread than that from temperature variations when imaging with my 130mm f/6 APO.
Since a 152mm ED doublet lens is about $2,000 and a 152 ED triplet is about $6,000, it made me think about the real benefit of the triplet when considering NB imaging.
It is likely a triplet does a better job of correcting field curvature and other off axis issues but enough to count it out?
With RGB imaging, the filter band passes are way bigger. Probably enough to show problems but in NB, maybe not. I guess all this could be tried with a smaller ED doublet refractor but I know there are optical engineers out there and was hoping for some feed back on the idea. Or even info and pix from someone already doing this.