The thing is, the "slightly dangerous thing" (i.e., uranium with all its associated breakdown products) occurs naturally in vast and totally uncontrolled ore bodies, sometimes right in contact with the water table. Yet somehow we're not worried about that (except in a very few cases where the concentration is extremely high). Why not?
It's like the way that no one gives a crap about the natural oil seeps off the coast of Southern California, which dump about 5 million gallons per year right into the ocean, then they have a panic reaction when a boat accident dumps a couple of hundred.
I dunno, maybe the localized increase over background levels? As they say, "dilution is the solution to pollution"... a little oil or radiation here and there will be reabsorbed by the ecosystem without much incident. Seemed like Fukushima was basically dumped into the sea and we didn't end up with Godzilla (too bad, really).
Likewise, some slow underwater seepage probably just gets washed away? But if you add a whole bunch of oil on a seashore that normally doesn't have much, it'll kill a bunch of cute animals and make your PR team have to work weekends.
If you drastically increase the background radiation near a suburb by piling spent fuel near it, the residents won't scientifically analyze the risk, they'll just try to keep you away. If you try to build a new plant, every politician within bribing distance of some special interest goes haywire.
Yucca Mountain might've been an option but that got shot down too.
A mix of precaution and NIMBYism, some of which is good science, some of which is just junk politics, and much of which is just people being primates. We're not evolutionarily equipped to make these sort of large scale long term decisions, only to live and die by our tiny, intimate lives.
It's like the way that no one gives a crap about the natural oil seeps off the coast of Southern California, which dump about 5 million gallons per year right into the ocean, then they have a panic reaction when a boat accident dumps a couple of hundred.