Great article here on the effects of Cherynobyl: http://www.who.int/ionizing_radiation/chernobyl/backgrounder/en/index.html
It does indeed say many thyroid cancers took years to turn up - but the number STARTED to increase right away, and it's fair to report that hasn't happened.
It seems unlikely that any effects of Fukushima will forever be impossible to count by keeping statistics. You cannot, even with Cherynobyl, ascribe a *particular* cancer case to the one cause, even there they can only say that "the cancer rate is higher by X%". They figure that some extra 4000 will die of cancer (than would have gone on to die of other causes later, of course) - but this is across hundreds of thousands exposed, so it's an increase in cancer rates of 3%-4% on that large group.
Chernobyl had the problem that they DIDN'T STOP DRINKING THE MILK in the area, the contaminated milk. Nobody made that mistake with any food near Fukishima. Worse yet, the kids in the area were iodine-deficient!
The cancer rate increase from Fukishima could be, say, a hundredth of Cherynobyl's (it's probably less), and be 0.03-0.04% ... you'll never be able to say whether the number is higher or not, because the error bar on just COUNTING cancer deaths (when Grandma has cancer and dies of a heart attack, would she have had the attack without the cancer? A doctor's call on that can change the outcome.) is much higher than 0.05%.
The cancer rate around Fukishima could be, say 100,000 dead out of 300,000 people when we add them all up 60 years from now - when the stats said it should have been 101,000. Then some stats guy will have to wearily explain that it was really 101,000 plus or minus 4,000 - and if only 100,000 died, then in that area's case it would have been 99,890, because by 2020, researchers using the disputed "no threshold" model had put the probable deaths at 150.
So our real story here, is why are we caring about a death rate that is smaller than a statistical error bar that nobody gives a crap about, at least as a news story.