My two cents: for a reasonably simple structure such as the "carbon
around swim tube" proposed, I think the best approach may be to simply
build one or two of them and test it/them to destruction. Compare this
against a likewise destructive test of the wood part you're proposing
to replace. Allow a generous margin for error -- especially if others
are going to be building it from your plans, because there can be a
great variance in composite strength depending on the builder's skill
and environmental conditions.

Greg