One spec for bipolar transistors frequently overlooked is the reverse-bias "leakage" current from collector to base. If you look at just those two terminals of the transistor, you have a reverse-biased diode. All diodes have some amount of reverse-bias leakage current. This spec for a 2SC2879 is really small. Primitive transistors from 60 years ago would have enough of this leakage current to keep you from shutting them off properly. When the industry changed from germanium transistors to silicon this problem became a lot more rare. A 2SC2879 is actually a dozen separate transistors all in parallel. If one of them breaks down between collector and base, the current coming out of that base terminal will 'turn on' the other eleven transistors in parallel with it.
And that also gets back to my suggestion of checking the bias resistance and bias voltage as the first step. The issue would have shown up then at the beginning, and then followup with testing the transistors.
When an Emitter and collector are shorted, the transistor cease to work as a switch. Basically, the electrons from Emitter to reach Collector by bias and hence it becomes switch. Now you have shorted, both ends of the switch its basic function as switch is defeated. Am I wrong?
