There are two cases to consider. Accident and normal waste. Accidents aren't nearly as contained. Normal waste is contained for now, but what about long term? We are looking at time length longer than any government has been stable.
>Two, some waste products decay alpha particles which are basically harmless unless consumed.
...some...
Also, what vectors could these particles enter into food when you consider their lifespan?
>Three, some waste products have short half-lives which increases their risk in the short term but substantially decreases their risk over a 500 year period.
But to what extent are these the ones being pushed for?
>Four, most deaths from nuclear waste would be slow and probabilistic (e.g. 18% increase cancer risk for an individual with high exposure) whereas deaths from some other power source would be immediate and devastating (e.g. flood, destruction of habitat).
Are we really suggesting that the death isn't comparable because it is a long term probabilistic? Imagine if some company dumping sludge into drinking water was making the same argument. Would we find it acceptable?