Devilsbrew, Yes, any insulating material which encases the mass will slow the transfer of heat, but i was planning to build a 2x2 stud wall an inch or so away from the mass to allow for airflow. My original build had wood in direct contact with the mass (8" system) with no safety problems, but I always prefer a little more fire safety.
Hagrid, here's a different design. I modified a drawing i'd done some time ago. Be aware the output from the drum should be on three sides of the burn tunnel so as not to constrict the flow, i just didn't draw it that way because it becomes too cluttered to easily comprehend.
If you were to cast a 5" riser with 2" thick walls, you could use a 14"-16" diameter 30-gallon drum instead of the pictured 50-gallon drum. A 5" riser has a CSA of 19.6", so your burn tunnel could be 4" square, 16 sq in. (A burn tunnel smaller than the CSA of the stove makes it more 'rockety'. The air just sucks through faster)
Note the Peterburg fresh air injection plate in the feed chamber, and the Peterburg 'tripwire' cast into the ceiling of the burn tunnel. The plate can be 1/8" thick metal, and there should be a 3/8" gap between it and the face of the feed chamber. It should curve slightly into the burn tunnel. The fast-moving air siphons in extra oxygen, improving the combustion efficiency. The tripwire (cast at an angle across the ceiling of the burn tunnel) makes the hot gases tumble, bringing more oxygen into contact with fuel particles. The majority of the combustion in the burn tunnel takes place along the ceiling, where it's hottest.
Darrell "Jake" Jacob, Oddmar on all the forums, KC9PZN to all you amateurs.