I am not sure what the other core devs will say. I'll try to give some background.
FLTK uses and links NanoSVG. So the same limits that apply to NanoSVG apply to FLTK's sig support as well. We do update from the original source regularly, and we do minor adaptations for compatibility. We do not plan to add features to NanoSVG. We try to keep FLTK fast and light.
There are three things that come spontaneously to my mind to work around this. The first one, just load a png or inline one, you already tried. If you want to stick with vector file, just link a different SVG interpreter. There are several libraries out there, and you can import the result using RGB. The third one will give mixed results, depending on the resolution of the target machine: use the commands in the SVG file and map them direct;y to FLTK graphics calls. This requires a little bit of coding, but FLTK rendering call include lines, loops and arbitrary filled areas, which is enough to draw any flag that does;t require shading.