Which one is more useful to mention in the manuscript either DOS plots or Band Structure to describe the electronic properties? Which one has more weight-age? Secondly, what properties could be address by DOS/PDOS and Band Structure?
The best way to answer your question is ask to yourself: what information i want to know? If you want to know about band gap energy, the better is the DOS plot. Otherwise, if you want to know more details, such as direct or indirect band transitions is better to use Band Structure plot.
Band-structures are only really useful for visualising the electronic structure, and seeing over what energy ranges excitations can occur. For everything else, including actually computing the probability of excitations, you want the DOS.
Essentially, band-structure plots are just a way to visualise what is actually 4D data -- how the band-energies vary across the 3D Brillouin zone. By choosing symmetry directions we know that the lines in the band-structure diagram represent either local maxima or local minima in the band-energies, so we can see where there are allowed and forbidden energy regions; but we cannot see how much of the Brillouin zone allows these transitions.
The DOS plot tells you directly how many states there are in any given energy window, and so can give you direct insight into the likelihood of any transitions. Of course, the true transition probability will depend on the transition matrix element, but the DOS is a good way to see *why* certain transitions are particularly likely or unlikely.