Okay, but I'd still be interested in:
a) A measure of death risk to the general public, i.e. people not directly involved in the production of the energy. It's important to distinguish internalized vs externalized risk.
b) A measure that accounts for the potential cost of preventing these deaths. That is, one that could capture insights like "source X and source Y currently have the same death count per kWh, but it would cost half has much to achieve a 10% reduction in deaths from X compared to Y".
The ratio of internal/external deaths would hugely effect the measure in b) -- it's much easier to stack on measures that can prevent deaths if they're employees of the provider than if it's the general public.
And so, on a), I would guess that measure for wind power is really hard to distinguish from zero.