What makes pollen varieties useful index fossils? Should pollen not be a poor index fossil?
Depends upon the species. If the pollen spores are large enough, and of wide spatial distribution, and easily recognizable, and preserves well in sediments, and of distinctive age range (geologically), then it could be used.
There have been mistakes in the past where modern pollen collected in microscopic pores in older rock, giving the illusion of rock ages much younger than they really are, so precautions against sample contamination are of paramount importance.
In addition to what @GordonStanger said, there are a few other points to consider.
- Pollens and spores are made out of sporopollenin which preserves way better than what one would expect.
- Although there isn't probably a perfect 1-to-1 correspondance between pollen morphospecies and the plant species that produced them, there is still a wide variety of pollen morphospecies, and they still have usable first and last occurrences.
- But the main point is that pollen are produced by land plants, meaning they have a continental origin. Yet, because some of them are transported by the wind or by water, you can still find them in marine sediments on top of finding them in terrestrial sediments, and this is a very valuable property: indeed marine sediments are easy to date precisely (thanks to magnetostratigraphy and the myriad of planktonic microfossil that sediment in them) while terrestrial sediments are really not (apart from pollen, small mammal remains can be used during the Cenozoic, but otherwise there really isn't any widespread biomarkers in terrestrial sediments, and geochronological dating is rarely possible). Pollen morphospecies first and last occurrences can thus be, in theory, dated with some accuracy thanks to marine sediments while being applicable to hard-to-date terrestrial sediments.