This guy is an idiot. Nuclear power can be accelerated and decelarated it just takes more time. We aren’t talking about a Ferarri here we are talking about something that can generate the output of a 1000 Ferarris. You use peaking plants for instantaneous unexpected changes.
You point to a interesting item in the original article, where the author is lying by switching conceptual domains. Current nuclear power plants are designed for base load (constant full power), but there is no inherent reason that future nuclear power plants can't be so designed, or that nuclear plants used other than in the power industry aren't so designed now (submarines). It's an economic question, not an limit in physics.
It was a matter of tradeoffs, back when. Base load nukes are extremely capital intensive, and pay off better when run constantly. Coal plants are less capital intensive, and more of the cost is the coal, so that they are relatively cheaper to cycle. Gas turbines are even less capital intensive, but use more expensive fuel, and are used for peaking. (Peaking need is when the load grows too fast for coal plants to ramp up to met it, or when load exceeds the total nuke + coal output).
The engineers did a study, and concluded that the most cost-effective mix was nukes for base load, and coal for cyclic load, and gas for peaking.
The economics have changed, and it might be time to design a cycling nuke plant. And it would be strange indeed if nuke subs didn't already use such a design.