In general, quasi or pseudo reference electrodes become a necessity when there are non aqueous systems, molten salts or high temperatures involved. When you use traditional aqueous reference electrodes in these electrolytes, you risk contaminating the electrolyte with water/filling solution solvent and also plugging up the reference electrode pores. Moreover, there will be a significant amount of reference electrode potential drift due to the high liquid junction potential across the frit interface in non aqueous systems. Hence, using a psuedo or quasi reference electrode calibrated with an internal standard will give you more reliable data.
For nano-electrochemical systems, one good advantage of using QRCEs would be miniaturization. It reduces the size, cost and complexity of the device you are trying to design. It also makes in-situ experiments easier to perform. Using a non-isolated Ag wire QCRE would be a much more simplified setup than say, having an Ag/AgCl reference electrode isolated from the bulk solution using a glass frit or salt bridge, and a counter electrode positioned far from the working electrode. You can even design a simpler 2-electrode cell which balances the working electrode with a single QRCE. (if the current measured is very small and the surface area of the QRCE is relatively large compared to the WE).
You can use a two-electrode cell instead a three-electrode potentiostatic system if the currents are low (as in the case of micro or nano electrodes). If you uses only two electrodes, the current passes between reference and working electrodes, but it is very small. Then, the ohmic drop (iR) will be also small and does not affect the potential (applied to the WE against the RE). In case you have higher currents, you use a third electrode, the auxiliary or counter electrode. In this way, the current flows between WE and CE, and the potential of WE (against RE) remains constant.
As already outlined it depends on your system. I do not think there is such thing as a leak free reference electrode, so if you perform measurements at low ionic strength in a small volume (compared to your reference electrode) this may cause difficulties. And sometimes reference electrodes are just impractical to use in microliter volumes. If your measurements are performed with a reversible redox couple you may find that for example a carbon fiber or Pt wire will assume an equilibrium potential close to the redox potential of this couple so the reference potential will be that of the redox couple (i.e. 0 on a voltammogram). Also as above you can do this in a 2 electrode system but the v-gram may be a bit flattened (spread over a larger potential interval), but if you are integrating charge this is no major problem but confirm the counter electrode capacitance does not become a current limiting factor. If necessary you can coat the counter electrode for example with a polymer having a high affinity for the redox couple used, so the size of this electrode can also be minimized (for low scan rates). Possibly a bit of confidence is lost with respect to the potential (most people tend to report the potential with respect to a familiar reference electrode, even if the actual potential of the ref. electrode is not thoroughly checked). However you may find that the benefits of a pseudo-reference (not such a good name), compensate for this when you consider noise and most of all the experimental conditions.
As you said the equilibrium potential of the quasi ref electrode is close to the redox couple potential. Is this true also for a molten salt at 200 C ? if not, how is possible to calculate the equilibrium pot. of the ref electrode in this case?
Nourhan Mohamed ,I have no experience with molten salt electrochemistry, and I guess a lot will depend on the system and it's composition. However several reference electrodes for molten salts exist, so it's possible to determine the potential of the quasi reference electrode under various conditions. If this is unsuitable you can consider an internal reference (i.e. a reversible redox couple which is stable under the experimental conditions e.g. Fe/Fe2+ generated in situ). Again a lot depends on the system. But if the system involves a reversible redox couple, then what is true for "ordinary" electrohemistry is also true for molten sals. But to calculate the true potential of the quasi reference electrode a known reference is needed. Just as the SHE was needed to determine the potential of calomel , Ag/AgCl etc...