Throwing some numbers out.
In 2018, nuclear accounted for 4% of global energy production [0]. Since we have been shifting away from nuclear, this represents an underestimation.
Going back to Chernobyl, we have had 2 class 7 nuclear incidents. Scaling this up 25x, this gives us about 1 class 7 nuclear incident every 8 months; assuming no advances in safety occurred since the 70s, when both plants were built. For simplicity, I will round this down to twice a year.
So, how bad are these class 7 disasters? According to WHO estimations [1], Fukushima caused an estimated 0 acute deaths from radiation; 400 long-term deaths from radiation and 1,600 deaths from evacuations (I suspect elevated due to the trigger tsunami). For reference, the trigger natural disaster had about 15,897 deaths [2].
So, in terms of casualties, Fukushima made a natural disaster ~10% worse.
In terms of cost, Fukushima is estimated to cost about $187 billion USD in decommisioning and compensation (Japan Ministry of Economy, Trade, and Industry, 2016) [3]. In addition, the disaster also caused an estimated $266 billion USD in fuel imports to replace the lost power (as of 2016, so 5.5 years of imports); I should discount the hypothetical operating cost of Fukushima, but can't find the numbers for that.
So, to get to 100% nuclear power, we are looking at 2 Fukushima's a year. That is 3,200 deaths and $374 billion USD a year. (In practice, this would not work because there are grid problems once nuclear becomes a significant chunk of power production, but we are really talking about the margin; just extrapolating out to make the numbers clearer). Also, these costs are concentrating in the near term to the incidents, so will go down fairly quickly once we shift away from nuclear.
I was going to compare this to global warming, but it is hard to find good estimates on the damage, and this comment is also already involving much more research then I intenended to do.
In 2010, an estimated 3,000 people died from fine particle pollution from US power plants. [4]. This already almost exceeds are estimate for annual global deaths from nuclear power, so I can stop looking for all the other deaths non-nuclear power causes.
Estimating the cost of not nuclear is more difficult, and at this point I already spent way too much time on this.
I would, however, like to reiterate that this analysis started with me (in my judgment) comically overestimating the cost of the counter-factual nuclear disasters.
[0] https://en.wikipedia.org/wiki/World_energy_consumption
[1] https://ourworldindata.org/what-was-the-death-toll-from-cher...
[2] https://en.wikipedia.org/wiki/2011_T%C5%8Dhoku_earthquake_an...
[3] https://www.wiseinternational.org/nuclear-monitor/836/econom...
[4] https://www.catf.us/educational/coal-plant-pollution/