If two boats are going to weather and one is pinching waaay too high and another is footed off to a reach it is still possible to plot their data, yes?

Though neither represents the optimal course, the best vmg will lie between the two, correct?

If so, it is then only necessary to have the two alter course gradually towards each other. One or the other must then sail the optimal course if only by accident.

Plotting large numbers of isolated data will eventually demonstrate the optimal course.