Rod Elliot, good information and a purveyor of some excellent audio amplifier designs.
I wish I had kept some of the thermal imaging pics I had saved. They show the path of heat dissipation close to the device junction area like no other explanation. If we only needed to dissipate the junction heat to half it's value, it could all be done with a heatsink just slightly bigger than the junction footprint (with active airflow). As it is, we need larger sinks so that the sink itself takes the heat and dissipates it within the mass of the sink.
I make some aluminum cases and the use of an all-aluminum case aids quite a bit with overall heat dissipation, not just the device junction temps. When you theorize about heat dissipation, it's good to keep in mind that sinking the high current devices is only part of the task. Critical, less durable components, also need their heat removed. I know in Mex's situation, he is not adverse to using adequate fans for airflow. That is a good thing.
