Part of the answer lies in what you plan to do after staining. If you just want to look at the cells, then you could try both and see what the results are. The big difference is that methanol will dehydrate the cells and they will be very flat, two-dimensional, while paraformaldehyde will fix them in three dimensions. That is much more useful for many procedures, such as confocal microscopy with identification of other structures labelled with other fluorophores. Permeabilization is not needed with Hoescht 33342. it stains live cells quite well. I have not tried that with DAPI, however.
Dapi is also cell permeable so you would not need to permeabilise.
For looking at protein localisation etc then paraformaldehyde is generally used with methanol fixation being used (in general) for RNA work. Paraformaldehyde is a "cross-linking" fixative so it "locks" proteins in place. Methanol is a "precipitative" fixative (RNA and proteins will precipitate out as the Methanol competes for the water) so protein immobilisation is not guaranteed (it will also flatten the cells as mentioned above).
If you do methanol fix then you generally don't have to permeabilise anyway - as the methanol will cause "tears" in the cell membrane so it is both a fixative and a permeabilisation agent.