Leaving aside the fact the deserts are essential biomes on their own and shouldn't be treated as garbage dumps, didn't Fukishima and Chernobyl teach us that putting nuclear power plants "way over there away from anything" doesn't really matter in the case of a serious accident?
The current level of burning coal equates to roughly 50 nuclear meltdowns a year because coal contains thorium. So in terms of "putting it anywhere," an in-place switch from coal to fission in a populated area is actually safer.
In addition, the article identifies outdated designs as the primary reason for the cost of nuclear plant rollout. Guess what? Modern plant designs are passively safe, particle physics (not electronics, nor mechanics, nor humans) prevents criticality events.
As measured by GHG emissions, nuclear power is certainly better for the climate. Whether or not that equates to "safer for human activity"? The devil is in the details.
If the reactor is passively safe, the amortized level of radioactivity is zero, especially if the waste is "burned" in secondary reactors (such as TWR).
Amortized over the duration where we are typically seeing meltdowns in gen1 reactors. 10-50 years?
> "if"
The article talks about modern designs. It is an aspiration, not an assumption. The central argument is that adopting modern reactor designs eliminates meltdown concerns (in addition to reducing costs as the article suggests). That isn't an assumption, that is a suggestion.
1. https://pubs.usgs.gov/circ/c1198/chapters/207-217_RadWaste.p...