The Jan 06 edition of Blade has a nice article on blade grinds. According to the article (which so happens to coincide with my own opinion) the best grind for most applications is a full flat grind with a convex edge. It combines a low profile with a strong edge. However, all that is pretty much meaningless (a Cliff would probably haste to point out) as long as you don't specify the parameters of the grinds, most importantly thickness behind the edge (were the edge grind transitions into the main blade) and edge bevel.
This edge geometry is pretty easily achieved. On any flat ground Spyderco model I like to start with a 12 deg bevel on the edge. Even though I am pretty good at holding an angle free hand I am naturally round over the edge bevel free hand anyways. Depending on the effect that I want to achieve, I raise, after polishing out the scratches of the previous stone, the angle on the finer stones (usually I progress 700, 3000, 10000) a little bit, this ensures that I work on the edge and naturally rounds the edge to a slight convex shape. I end on the finest stone at about 15 deg. The final stropping with CrO increases the angle naturally ever so slightly further.
I have reduced the thickness behind the edge on my Pacific Salt a little bit more as an experiment than a necessity, but usually I don't adjust the thickness behind the edge. I find that Spyderco does a decent job of adjusting the thickness behind the edge depending on their model and its intended task. The Lil'T for example is pretty thick behind the edge (about 0.03 in) while the Caly Jr. is about half that, which is in keeping with the indended usage of the two models I would say (the Caly Jr. is a performance slicer, the Lil'T is a brute). Adjusting the thickness behind the edge is not easy on a flat ground model (much easier on a hollow grind, a lot less material to remove) without a motorized grinder. Still possible with a good coarse stone though, but it takes a lot of elbow grease and will scratch up the blade a lot.
There is something to be said for a full convex grind or even better a median ridge geometry. I wish Ed Schempp would say something about the latter since his fixed Camp blade will feature this grind and the arguments sound very exciting. But basically a full convex grind will limit the contact of the material to be cut with the blade if the cut is deep. This reduces potential binding. The median ridge geometry has the same advantage but with the additional plus of a more balanced weight distribution between edge and spine. Obviously an argument that is irrelevant for a small slicer but important for large blades like a Katana (which all sport a median ridge geometry).
I don't really see any advantages in a hollow grind except for the ease of adjusting the thickness behind the edge, but especially for deep cuts I would say that the hollow grind is at a clear disadvantage compared to a flat grind. At shallow cuts I doubt it will make a difference.
Hope that helps. As always, just IMHO of course.