It is tempting to think that the billion-star Gaia survey will be the final word on Galactic studies. However, the effective spectroscopic magnitude limit of Gaia is only V < 17, equivalent to a G dwarf at a distance of 6 kpc in the absence of dust. We can go much deeper than this with large-ground based telescopes. Such observations will be essential if we are to extract chemical information over a large volume of the Galaxy. Future ground-based optical and infrared surveys will likely map the Galactic bulge and provide a self-consistent model for the bulge and bar. Deep pencil beam surveys towards dust-free inner windows will establish the strength and orientation of the spiral arms, and allow us to correct the rotation curve (and Oort's constants) for streaming motions.
A self-consistent dynamic model of the Galaxy is essential in understanding the smooth underlying potential. Once we have established the basic parameters reliably, we can revisit the distribution of dark matter within the Solar Circle and beyond, an essential piece of information if we are to reproduce the main features of the Galaxy from ΛCDM simulations. The kinematics of the outer halo stars will ultimately constrain the shape and figure rotation of the dark matter halo.
Binney (2005) has highlighted the need to get ready for the impending data revolution. Now that Gaia is fully funded, the case is even more compelling. Binney makes clear that we need a consistent multi-component model that can be modified to give a better fit as data become available. To complicate things further, we suspect that these models may need to take on board the fact that the Galaxy is not in dynamical equilibrium, and therefore to think in the wider context of the Local Group. This may become possible with improved 3D space motions of individual galaxies (Brunthaler et al 2006). Finally, Binney's vision considers only the phase space information when in fact the chemical space is likely to be equally unwieldy (Bland-Hawthorn & Freeman 2004).