Microemulsions are supposed to be turbid and nanoemulsions are supposed to be transparent. So should I consider the point at which turbidity occurs as end point?
The usual definition is the opposite of what you state: microemulsions are transparent, with smaller drops than nanoemulsions, which are more turbid. Nanoemulsions can be almost transparent, depending on the drop size and refractive indices of oil and water phases.
For me, the big difference between the two is that microemulsions are thermodynamically stable, so they form spontaneously, whereas nanoemulsions are not, so you need to use a high pressure homogenizer (or similar) to produce the small drops. Therefore, phase behavior is much less important for nanoemulsions. You just need a good emulsifier, a good homogenizer and a formulation that prevents Ostwald ripening.
Indeed as Prof. Alan Parker said, microemulsions are clear, transparent, thermodynamically stable isotropic dispersions of water, oil, surfactant and cosurfactant, formed by self-assembly. In turn, nanoemulsions are translucent, slightly turbid or opaque thermodynamically and kinetically stable isotropic systems, formed by high mechanical shear.
Let me just emphasize the point because it leads to much confusion: A phase diagram is not an ordinary diagram. It is a map where the thermodynamically stable phases of a system are shown for a set of conditions. It conveys real thermodynamic information.
Microemulsions are thermodynamically-stable systems. Nanoemulsions are NOT thermodynamically stable. Therefore, the term "phase diagram" only makes sense to microemulsions.
What you do for nanoemulsions is then to just map the processing conditions to see what works best to form the nanoemulsions. It may look like a phase diagram, but since the nanoemulsion will depend on the sample history, it is not.