This Nature article dubs TRISO fuel as "accident-tolerant", which is just as well given the lamentable history of it in actual use.

https://en.wikipedia.org/wiki/AVR_reactor for instance, which suffered so many accidents that it was known as the "Shipwreck". It ended up with TRISO pebbles getting stuck in the reactor vessel, the primary circuit being hopelessly contaminated with fission products, and is now impossible to safely decommission using current technology.

It seems to me working SMRs really would be valuable, particularly for off-grid applications or where process heat is needed alongside electricity generation.

But as someone who was, like you, taken in by the pebble bed hype last time around, I've now learned to be much more sceptical about new nuclear technologies until they've actually been proven in practice.