If the sizes of your slit shaped fiber bundle are within the input slit of the monochromator, and the angle of the beam from the bundle is smaller than the angle admitted by the monochromator, then external optics cannot increase the throughput. (Still, you might want to use it, e.g. for better spectral resolution if only a small part of the grating is illuminated without additional elements.)

If, of the size and angle as described above, one is bigger and the other is smaller, then most probably it's possible increase the throughput with additional optical elements. In other cases, everything depends on actual distribution of light brightness by angles and within the bundle.

One more thing to check might be light coming in cladding or at angles bigger than fiber normally transmits. For bundles, though, it is usually very weak, because a glued part of a bundle is often design to suppress this light. And for a short single fiber it can be considerable, and you might want to use or to suppress it, depending on your task.

And, of course, take into account the parameters of your detectors after the monochromator.