Polarisation modulation would have at least the same spectral occupancy
as amplitude modulation, as it is equivalent to amplitude modulating two
orthogonally polarised carriers. (You are actually modulating with the
sine of the signal, so it might actually be broader than AM and more like
narrow band FM. For the simplest analysis, the trig functions distort
the signal, inserting harmonics which would broaden a band limit signal.)
The reason we use systems which are broader than amplitude modulation
is that they have a higher error free bit rate for a given signal to
noise power density ratio and that they tolerate multipath reception
> average radio telescope, do they monitor polarization?
It would certainly be polarisation sensitive, although it would simply
convert the polarisation to amplitude. I seem to remember though, that
some radio astronomy objects do exhibit polarisation, so controlled
polarisation might be useful, it might also help reject local sources.