It's basic physics. Consider what happens (deliberately) at the active electrode. All that "pin-pointed" energy (high-density RF current) has to be safely dispersed over a relatively large area when leaving the patient (completing the circuit). In other words, low current density if any "heating effect" (burn) is to be avoided. I seem to remember at one time the "return plate" was often known as the "dispersive electrode".
But even the best "return plate" in the world won't be much good if it's not firmly in contact with the patient! It has to complete the circuit.
Yes, over the years various attempts at "return (electrode) plate monitoring" have been tried. Some have been more effective than others, but any ESU should have some sort of REM (unless, of course, it is truly ancient). And, naturally, it should be checked as part of regular I/PM!

Meanwhile, and in passing, it may be worth recalling that current will always take the path of least resistance ... so if the patient's leg is flopped against the bare metal of the table (for example), then that is where the "burns" are most likely to occur (asssuming, of course, that the table itself is earthed)!