PL works nicely for MoS2 where the monolayer has a direct bandgap while bilayer and thicker have indirect bandgap, so for monolayer has a huge PL signal compared to thicker flakes. I guess if you know how the band structure changes with thickness PL can help also in other cases.
In Raman you see the peaks shifting when you thin the crystal. I think you have to do at least once a calibration of thickness - Raman peak position. Measure the thickness by AFM (be careful it may overestimate the thickness) and Raman on the same flake. I'm not sure that only the Raman spectra is enough to tell the thickness as it may depend also on the substrate (e.g MoS2 on gold and SiO2 has bit different position of the peaks).