I agree. At least in my limited experience, it seems a lot of the deaths in the US are related to either intoxication or sleeping. In these states, it's common for the person to have never even been aware of the danger.
Since pushing hard for fire alarms and mandatory retrofit obligation, Germany was able to reduce fire deaths by almost 50% (study in German):
https://www.hekatron.de/fileadmin/user_upload/Studie_Wirksam...
Better homeless shelter policies & mental health support would be wildly more effective at preventing fire deaths than code enforcement.
That won't work as evidence that US fire policies aren't helping. You need data from somewhere where those policies aren't in place, like presumably Finland.
Policies that would help save lives are sprinklers in all houses. This policy is controversial due to the cost and maintenance involved. But the very scenario that has brought this about was a Maryland politician's son burnt down his house and died while passed out drunk. Now that state has a law requiring sprinklers.
No, you didn't identify any logical support. If the American policy ensures that people who are aware of the danger always survive, then (a) that is what the policy is supposed to do, and (b) it will guarantee that American fire deaths all happen to people who were not aware of the danger.
Given that, if the policy worked as advertised, it would guarantee exactly the observation that you see, it cannot be the case that this observation is evidence against the effectiveness of the policy.
The data you are looking for literally does not exist. That’s not something people closely monitor.