For example if nuclear costs 1bn for a GW and solar+wind costs 0.6bn but you need to also install 0.5bn of storage to make the grid reliable to the safety margin you deem acceptable then it is no longer a cheaper system.
Amount of storage needed scales up rapidly as you get towards an grid with entirely intermittent generation.
Storage is more expensive than generation. Always.
We see some storage on the grid because storage is fast. Batteries can respond to load changes a lot faster than anything else. Demand goes up, the batteries are drawn down while cheaper generators are spun up. A few places use pumped hydro for the same purpose, but that's very situational.
The problem lies with the standard of As Low As Reasonably Achievable. If you had reasonably priced nuclear power you could add more safety. Thus you have to add more safety. Thus you no longer have reasonably priced nuclear power.
And, when it comes to power it's a totally wrong standard. You should not be looking to minimize individual technology risk, you should be looking to minimize total risk of competing technologies. Making your nuclear plant safer instead drives generation to natural gas (~10x as dangerous as nuclear, plus climate effects) and oil (~10x as dangerous as natural gas, even more climate effects.) Or even coal (~10x as dangerous as oil, even more climate effects plus nasty stuff going up the stack--including emitting more radiation than would be tolerated for a nuclear plant.)