I have heard the same as Dick Hanson, that in some cases you can actually see a drag REDUCTION with a properly truncated trailing edge. The idea is that the low pressure in the separation bubble behind the trailing edge will reduce the adverse pressure gradient seen by the boundary layer it approaches the trailing edge, thus keeping it attached. Unfortunately, I don't know of any "rules of thumb" or analysis tools that might help here.

I would expect that if you didn't have a boundary layer that is at or close to separation, truncating the trailing edge would result in a drag penalty.