The problem with these kind of statistics is that mere numbers mean very little, because the sources of the deaths are very different. Do you count roof work? Do you count mining? Do you count the creation of construction materials and their mining? How do you count cancer from slightly increased nuclear radiation? Especially when it's hard to prove whether it really comes from nuclear waste, but it certainly could be? And are birth defects deaths? Are miscarriages?

These are things that need to be addressed for these kind of statistics, or you end up comparing apples to oranges.

Still, I think everybody can agree that the death rates for coal and oil are insane. However you compare nuclear vs solar, coal and oil need to go.