The title of your question indicates it is a binary organic working fluid.
PTC 46 specifically indicates steam as the only permitted working fluid. In the case of a binary working fluid ( organic or not) I think one would need to ascertain the amount of each constituent present in all parts of the cycle , from start to finish of the test, and account for changes across the system boundary as well as changes in the fluid holdup within the system boundary.
I'm not totally familiar with the thermodynamics of mixtures, but there may be some way of "cheating" the test by having the end state of all fluids at the end of the test have a greater entropy than the starting state. If so, there needs to be a means of limiting this method of "cheating".