Hi Jeremie,
it sounds that your Problem is more a 2D - problem than a 1D - problem since flow seems to be distributed widely in a vast foreland.
Therefore I recommend to use the 2D - Software MIKE21FM. If necessary, MIKE11 - branches can be coupled to the MIKE21FM - it is then called MIKE FLOOD. The coupling should be done, if you have many discrete branches in your model and the floodplain is flooded with well known most likely flow paths.
But, if you can numerically effort resolving your branches in an unstructured mesh, please use the adequate software.
Everything else is an extreme simplification of the natural conditions an should be avoided if possible.
Bests Christian