Faint magnitude limit for gen_SimObs_from_sedgrid too bright

[cohenrogere]

When using gen_SimObs_from_sedgrid to generate synthetic input and output photometry, stars with input magnitudes near the detection limit (or faintward) are not generated. This is crucial in order to use simulate_obs to generate additional "artificial star test results" for which the input and output magnitudes are known. In addition (and maybe relatedly), the gen_SimObs_from_sedgrid optional parameters compl_filter and magcut do not appear to be functional, although the weight_to_use parameter does function correctly.