I once started a bugfixing week with fixing a bug that caused around 1% of the crashes. My team lead didn't think that was either smart, obvious or very anythingatall, and said so loudly and clearly in the team meeting.
Citation needed.
These plants are running now. Are you saying the electricity they are generating is not being used? The question is what replaces those joules if they shut them down?
See https://www.haustec.de/heizung/waermeerzeugung/waermemarkt-w... for example. There's also the chemical industry, which uses gas as chemicals, not as a source of joules.
The plants are running now, but that doesn't make them able to replace gas. Most of the gas can't be replaced by them for the reasons in my other response, but I think your question was about the rest?
The rest is because of the different traits of nuclear and gas-fired power plants. Nuclear power plants aren't adjustable, they provide the same amount of power, day and night, plodding on, with only very slow adjustments. Gas is the opposite. Gas-fired power plants are easy to adjust on very short notice, to match the unexpected changes in demand. If you try to replace quick-changing power plants with slow-changing, the network's frequency (50Hz in Germany) suffers. When demand shrinks and the slow-changing power plant can't adapt, the frequency rises above 50Hz. When demand rises quickly, the opposite happens.
Some adjustment is possible by involving other entities than just the two, so it's not quite 0%.
Nuclear is not really a replacement for gas, rather than coal, due to its slow nature. Gas power plants are used in Germany mostly for quick peak power.
If you fix a bug that currently causes few crashes but would later on turn out to cause big issues, and you see that outcome in advance, then it would in fact be smart wouldn't it?