The last study I used in development of a pilot ground school was from
2002 UK CAA White Paper. This is very likely the source of the information you have on hand, given the paper did rely on previous incidents and accidents in its analysis. The paper contains several charts and tables outlining the probability of survival for a hidden fire. It's been a few years since I looked at the data, but 15 to 20 minutes sticks out in my mind as a 50% probability of survival and anything after I believe it was 40 minutes was non-survivable, but I'm willing to be corrected if someone digs deep into that paper. As I said, it's been a few years and it may have been updated since I last read through it. I suppose the good news is that if the tables have not been updated, it could mean there have not been enough incidents to warrant another study. Or, as you infer, better design and components have increased the probability of survival.