>>A house can be thirty feet from an entire forest on fire and never burn down. We learned this through hardcore experiments in Canada where they built homes to test and lit forests on fire nearby.
Fires in CA apparently went as high as 300 feet (~100m) so no, that would not have helped. Heat alone, with zero wind would probably melt everything with xx meters. If they want to build, don't build in forests or clean out a zone and build a 50-100m buffer. If it was not burned today, it will be burned tomorrow.
Air is a ridiculously bad thermal conductor - it's thermal conductivity is 0.024 W/mK, so to deliver the same heat flux as a 100W light bulb through a meter of still air would require a temperature difference of about 4000 K. Most heat transfer through air occurs through convection (wind). Even then, if you can keep a 5 foot buffer zone of still air between combustible material and the raging inferno, you're fine. This is how you see pictures of scientists standing 5 feet away from a moving lava flow.
As the article states, though, it really is all about the embers. If you keep burning material from landing on other combustible material, fire is not going to start. That's why it suggests using non-combustible materials for roofing, avoiding areas where embers can catch, and keeping treetops away from structures.
While you are right that conduction is not really a problem, and right that embers are really important, it's worth mentioning that radiation from a fire can also be a significant source of ignition. And a large "flame front" can ignite structures from surprisingly large distances!
I haven't read whole article yet, but here's an abstract:
Residential losses associated with wildland fires have become a serious international fire protection problem. The radiant heat flux from burning vegetation adjacent to a structure is a principal ignition factor. A thermal radiation and ignition model estimated structure ignition potential using designated flame characteristics (inferred from various types and densities of vegetation) and flame- to-structure distances. Model results indicate that ignitions from flame radiation are unlikely to occur from burning vegetation beyond 40 meters of a structure. Thinning vegetation within 40 meters has a significant ignition mitigation effect.
width or not means that there's a LOT of material to be burned and to collapse into the next lot. I've burned wheat fields after 90% of straw was removed and I can tell you that in certain days it rises to 10 feet and you cannot stay close to it, it's unbearable. Not necessarily flames but heat, that barely visible thing. A week later you see trees 20 feet away with burned leaves. The other thing you learn when you clean canals and lots is that fire is unpredictable. Very.
Now imagine real forests with underbrush and 150 feet tall trees.
Fires in CA apparently went as high as 300 feet (~100m) so no, that would not have helped. Heat alone, with zero wind would probably melt everything with xx meters. If they want to build, don't build in forests or clean out a zone and build a 50-100m buffer. If it was not burned today, it will be burned tomorrow.