Dear All As Jaime said, the only drawback is the bigger memory and thus time requirements. And as you see in his link and former literature, bands can be perfectly plotted.
I just want to point to you all to this reference which may be interesting in your case, where all the calculations were performed with SIESTA trying to match previous a previous ABINIT version of the calculations. While in the final paper we didn't do all the final calculations with SO, we also tested the SO in the bulk and monolayer, but it didn't affect what we were looking for in the initial calculations of the NWs in the cases we were interested, and as Prof. Ferrer said it makes the calculations slower, so we proceeded just with spin polarized. In any case, if you have problems reproducing literature I can look up the psf files we used for the MoS2. https://www.researchgate.net/publication/257247197_Atomic_and_electronic_properties_of_quasi-one-dimensional_MoS2_nanowires All the best
