I know that galaxies can eject gas due to supernovae, accreting black holes, etc. These galactic "winds/outflows" are often defined/detected observationally using blueshifted absorption lines in spectra. But why is it always blueshifted absorption -- what about redshifted absorption? Does it have something to do with the outflow velocities being measured relative to the galaxy central/systemic velocity, such that redshifted absorption implies inflow instead of outflow? But what's the geometrical picture?
Yes, a blueshift here would mean relative to the systemic velocity of the galaxy - which means that, relative to the galaxy, the material is moving at least partly in the direction of Earth.
In many cases the outflow may be a pair of jets being ejected in opposite directions out of the nucleus of the galaxy, but we are less likely to see absorption lines from the jet on the far side of the galaxy (relative to Earth) because the disk of the galaxy may be blocking our view of it, and in any case it won't produce absorption lines because it isn't between us and the rest of the galaxy.