[Nathan Slingerland]

That seems like a reasonable way to think about it. And by that logic maybe we can 'derate' the power rating of any resistor stuck inside a sleeve like this one as 1/3 of its normal free-air rating.

I wonder though if we need to even assume 300 C as the reference since solder melts at 183 C. It seems unlikely that any of our resistors actually ever run - even at the hottest point - at that temperature.