Article fails to mention one small Army reactor (3MW rating) - the one at the (remote) Idaho National Laboratory. In Jan. 1961, for 4 milliseconds it generated 20GW ... then it killed all three military operators ... then released 1100 curies of fission products into the air.

A part of small reactor history often dis-remembered by people who claim 'no deaths'.

[https://en.wikipedia.org/wiki/SL-1]

The most significant information in this is the addendum where they discussed the changes to costs:

Solar and wind DRAMATICALLY fell in just three years (written 2015, updated figures from 2018 Lazard), which is the real challenge to any nuclear project: Already at an economic disadvantage, a nuclear project needs a decade to complete often, and while cost improvements to solar and wind will likely flatten at some point, it probably has 10 years of year-on-year 5-10% cost improvement ahead of it, and grid batteries are likely going to be even more dramatic in cost improvement.

"Worse, attempts to make them cheaper might end up exacerbating nuclear power's other problems: production of long-lived radioactive waste, linkage with nuclear weapons, and the occasional catastrophic accident."

All of which are fundamentally addressed by MSR/LFTRs if only they could work. Small scale may be the only way that LFTRs work because the difficulties with molten salts may require constant refabrication of the parts. At least the fuel is liquid, you could just pipe the fluid to the replacement plant.

ORNL's plant was small, but maybe its operation/scale was too small, or didn't actually perform in utility generation, so it didn't get included here.

Anyway, allegedly China is starting up a MSR planet next year.

And is it true that MSRs are forbidden from research by the federal grant foundations? That was a claim in some LFTR videos.