I'm no arborist Diablo (in fact my tree-felling methods have disgusted members here in the past) but I don't think it will be easy to produce a design for a shaft that long, that runs stably. The shaft will be subject to a phenomenon called "whirling": the middle of the span inevitably runs just slightly out of true and centrifugal force causes it to run further out of true, which in turn increases the centrifugal force, and it breaks catastrophically. The only solution is to have intermediate bearings along the shaft, inside the outer tube, to support it. You could use a substantial steel outer tube, and locate nylon bearings along it every 3 feet, and the problem would hopefully be overcome, but to be safe it would need to be bench-tested before running it under load and standing near it.
You were probably saved from "whirling failure" of the shaft up to now, by the joiner you had in the middle of the shaft. Chances are the joiner was not all that loose in the outer tube. If you could inspect the interior of the tube you'd probably find signs of contact by the joiner.
