Same compromise I made when I bought the base range version of my car with LFP chemistry. But I would not go lower in range than that. LFP is already much safer than any gasoline engine. I would like sodium just for the reliable range on low temperatures. Probably in the next years we will reach comparable density for sodium.

Battery density is energy per kilos. The problem is not only were to put the battery, but also the added weight.

Nuclear cannot manage fast transient, but for that we have gas peaker and batteries. But nuclear can indeed work in load following mode, with most modern nuclear power plant being able to reduce the amount of power significantly and circle during the day. The French fleet, for example is required to cycle between 20% and 100% twice a day, within 30 minutes. Modern reactors ramps up at 5% each minute.

That means that they can account for changes in demand. More data here: https://www.nice-future.org/docs/nicefuturelibraries/default-document-library/france.pdf

I have never disputed that in general solar+storage is cheaper, I am disputing the data in that Wikipedia article that make it looks like it is 20 times cheaper. It is not that much cheaper, and china build lot of nuclear because grid diversification is more valuable then just making it cheaper. Production cost and energy price are independent variables and nuclear bring energy price down as it stabilizes the grid.

Storage cost is going down, but storage demand by energy produced is going up as you need much more storage then just peak hour demand as you are shutting down load following power plant generator like coal, nuclear and gas.

The link I shared is to provide the reference to $62/MWh stated above.

The problem is that LCOE is an imperfect metrics that does not take into account storage properly for grid with high percentage of renewables (that requires significantly more battery storage than current 4h window considered in LCOE). LCOE also does not account completely for time effects associated with matching electricity production to demand. There is no clear metric for this, given that the cost depends on the structure of the grid itself and is specific for each country, but the Wikipedia article you posted show in the graph a very incorrect picture. Renewable (solar and wind) + storage is in the $80–150/MWh range, while nuclear is $130–200+/MWh range. It is worth noticing that nuclear cost is very high in Europe and US but can be actually very cheap (reason why china, the world leader on renewable is also world leader on new power plants). Estimation for new Chinese nuclear is at $62/MWh (https://www.renewable-ei.org/en/activities/column/REupdate/20240927.php)

As an owner of both an n150 minipc with 16gb of ram and an m1 8g air I can assure you this statement is false. The first is a toy compared to the latter. You can use it as inexpensive home server, but not as a work machine.