gstlal_inspiral_injection_template_match: fix chi computation

At line 50 chi_eff was computed using the spin magnitude (without sign) rather than the z component. I believe this is not consistent with the usual definition of chi. This merge request aims to fix this by using the z components for the computation of chi_eff. If this was intentional, please ignore this merge request. However, one should make sure that also the template chi (sngl.chi at line 39) is consistent with this.