The efficiency of early mixing is important for the interpretation of stars in our own galaxy that have ultra-low metallicity - lower than the mean metallicity of 10-2-10-3 times solar that is likely to have been generated in association with the UV background at z > 5. For a comprehensive review of what is known about such stars, see Beers (1999). If the heavy elements were efficiently mixed, then these stars would themselves need to have formed before galaxies were assembled. The mixing, however, is unlikely to operate on scales as large as a protogalaxy - if it did, the requisite bulk flow speeds would be so large that they would completely change the way in which galaxies assembled, and would certainly need to be incorporated in simulations of the Lyman alpha forest.
As White and Springel (1999) have recently emphasised, it is important to distinguish between the first stars and the most metal-poor stars. The former would form in high-sigma peaks that would be correlated owing to biasing, and which would preferentially lie within overdensities on galactic scales. These stars would therefore now be found within galactic bulges. However, most of the metal-poor stars could form later. They would now be in the halos, of galaxies, though they would not have such an extended distribution as the dark matter. This is because they would form in subgalaxies that would tend, during the subsequent mergers, to sink via dynamical friction towards the centres of the merged systems. There would nevertheless be a correlation between metallicity, age and kinematics within the Galactic Halo. This is a project where NGST could be crucial, especially if it allowed detection of halo stars in other nearby galaxies.
The number of such stars depends on the IMF. If this were flat, there would be fewer low-mass stars formed concurrently with those that produced the UV background. If, on the other hand, the IMF were initially steep, there could in principle be a lot of very low mass (MACHO) objects produced at high redshift, many of which would end up in the halos of galaxies like our own.