Psychological and social effects aside, I suspect most posters are overestimating the physical damage from a successful terrorist detonation in Manhatten. The terrorising effects of any such attack would be so great that the organisers aren’t going to waste effort optimising the physical effects. Hiroshima and Nagasaki can be used as a baseline, because in both cases the bombs used the minimum amount of material, with a “safety” margin for error. (The critical holdup in both cases was producing the material, so the pressure was to use no more than needed.) Smaller weapons can be designed, but such techniques will be beyond terrorist beginners.
But the areas of damage in 1945 were smaller than photographs can suggest. Judging by the maps in Rhodes’ The Making of the Atomic Bomb (p729 and p741), the serious physical damage was restricted to areas about a mile across. The damage for Nagasaki was particularly limited because of the sloppy aiming. Rhodes also has a plot of mortality against distance (p746) for Hiroshima. Even at 1 km, you had about a 30% chance of surviving. At 6 miles it was up to something like 99%.
There are then the two factors working in your favour: buildings in Manhatten are more substantial than in wartime Japan (but with frightening amounts of glass) and it won’t be an airblast. In 1945 the detonation height was optimised to maximise casualties. Terrorists probably don’t have that option.
If one’s interested in a crude period visualisation, Rhodes’ Dark Sun reproduces a 1950’s illustration (no. 75) of a Nagasaki sized bomb’s mushroom cloud against the New York skyline.
Dirty bombs are another matter, but personally I was struck by how low the estimates were in the recent BBC Horizon documentary. While intended to scare the bejesus of us, in their scenario (it goes off in Trafalger Square with the wind blowing to the east) even 6 miles downwind was only increasing your chances of contracting cancer from background radiation by about 1%. It gets nastier within a few hundred yards of course.
What they didn’t take into account is screw-ups by the designers. There’s a fair chance that, rather than dispersing the radioactive material, the conventional explosion merely sprays it across the facade of the National Gallery.
Finally, and to avoid sounding too sanguine about the whole thing, Robert Jay Lifton’s Death in Life (1967) is the classic sobering study of just how perminent the psychological effects are on the survivors.