I have seen far more PS-2 3V charging failures then I have seen PS-2 5V charging failures. There is a lot of old wives tales about various failures. MTH never did failure analysis I was told. Just anecdotal. Plenty of engines using BCR without engine failures. I have also received plenty of PS-1 and 2 with white batteries that still worked. Was the board at design limits? YES. Did the design use components requiring higher power needs? Yes. Just watch a startup on Z-4000. Jumps between 1 and 2 amps. Sometimes higher especially to drive audio.
So my experience is 9V batteries last about 4-5 years. Don't store well. Do gas out at terminal occasionally. Really rare to leak. So why. 7 small 1.2V cells with limited chemical energy in each cell. Used in series. So once one cell fails it kills overall voltage. Plus the board needs 5V for processor already drawing more current then later designs, so once a battery tips over in voltage harder to keep up. Plus the voltage is dropped to support the board.
So 3V the board uses much less current with modern chips. Has a 3.3V processor. The battery is now a 2 cell 2.4V battery with more chemical energy per cell. They seem to last 7-9 years with moderate use. Store well, don't seem to gas out or leak. 2 versus 7 cells. Plus board design has boost circuit that provides 2.5V via an inductor at shut off independent of the battery. So battery is 2.5, inductor 2.5 for 5V. Even when battery gets down low you still have power for the power supply (well designed 5V TI version) to feed the 3.3 V power supply for the processor. Plus board design kills audio volume to reduce load when voltage dips too low. Plenty of self protection features.
The real killer of the PS-2 3V is high voltage spike and shorts in the DC wiring (pinches, operator error). For the 5V age and component break down. Some die without even a short. Just won't power up anymore?
