New reactor designs, even with closely regulated designs, are very unproven. The next gen French EDF design recently had some units come online, but they had to figure out some unexpected buildup in some byproducts. You are also validating more than the design, you are validating a manufacturing materials control in a very long term feedback cycle - which is generally terribly difficult for making improvements.

In comparison to something like grid storage where one can site, build and operate, in less time than the make a minor nuclear design correction. Grid storage can correct mistakes at very low cost and on a very rapid cycle compared to nuclear.

Grid storage is proven through dozens of working installations earning decent money for their operators, pushing gas peaker plants out of the market.

Newer reactor designs are well simulated but have zero real world implementation.

Grid storage is as simple as E = mgh. We do that industrially, billions of times a day, worldwide, and have for centuries.

To cite uncertainty about storage is only to admit you have no argument.