Fitting the positions for diffraction peaks is an absolutely necessary step in the process for obtaining an accurate structural model. There are only a few challenges to do this. The first is starting with lattice parameters that are close enough to correct to provide significant peak overlap before refining lattice constants. The second is that one must select the appropriate corrections for the experimental artifacts that can offset peak positions to be refined, but this is done only after a good fit is obtained. Note that it is not sufficient if only one zone of reflections overlaps. For example, the 0k0 reflections only provide information about the length of the b axis, which is not sufficient to define the axes lengths even in a tetragonal cell. If n is the number of unique axes in a cell, for example 2 for hexagonal and 4 for monoclinic, one needs at least n and preferably 2n reflections where h, k and l each have two different values, such as (001), (101) and (030) are all different to be overlapped between the observed and computed patterns, so that that every lattice constant length is defined. High order reflections may not align as well as low Q reflections. It is usually fine to start refining lattice constants even if the high Q reflections are not overlapped, as long as for each axis there are some low order reflections are at least partly aligned.
In GSAS-II there are several ways to visually confirm that lattice parameters generate peak positions that do a reasonable job of overlapping with the observed peaks in the pattern. For all three, I will assume that you have both a histogram and a phase already imported into a GSAS-I project.