I don't understand how current is unlimited when that capacitor is not directly connected to the battery. I may look into a series resistor for higher voltages.

Also, I think from now on, for the VU meter transistors, I'm gonna put in a diode (cathode to base, anode to ground) at the points where -6V is reached. Then again, that would only be necessary if I tried 14V again.

I just ran a test with 7.2V supply for say 20 minutes and I didn't notice any heat, and I had the input volume cranked to about 80%. Distortion happened after that. I guess that's the price I have to pay for low voltage, but that's less expensive than overheating/damaging parts.

 

The circuit will work at 14.2 volts if you will follow my suggestions.
Run the simulation you have nothing to lose.

 

Ok so I did that and if I replaced a diode with a resistor, the Q2 current is almost the same but the power transistor (tip4x) emitter currents are much higher. Aren't diodes better than resistors in this case. I mean, it is sad if i have to draw 1.2A standby current as opposed to 318mA.

 

R13 should start at 100 ohms not 100K.

 

I don't understand how current is unlimited when that capacitor is not directly connected to the battery. I may look into a series resistor for higher voltages.

The capacitor C1 is a dead short at the beginning of a positive or negative audio signal from the output transistors when feeding the base-emitter diode of Q5. Then its massive current fed into the base of Q5 will probably destroy Q5. A series resistor is needed to limit the capacitor charging current into the base of Q5.

Also, I think from now on, for the VU meter transistors, I'm gonna put in a diode (cathode to base, anode to ground) at the points where -6V is reached. Then again, that would only be necessary if I tried 14V again.

Your added diode would also cause a massive negative charging current in it and in C1 unless there is a series resistor feeding it and the base of Q5, to limit the capacitor's charging current.

I just ran a test with 7.2V supply for say 20 minutes and I didn't notice any heat, and I had the input volume cranked to about 80%. Distortion happened after that. I guess that's the price I have to pay for low voltage, but that's less expensive than overheating/damaging parts.

With a 7.2V supply, the amplifier's max output when it begins clipping is about 6Vp-p producing 0.56W into 8 ohms.
Then each output transistor heats with only 0.28W when playing a continuous tone, much less when playing music or vocals.

The bias diode you replaced with a resistor had 18mA in it and about 0.75V across it then its "resistance" was 0.75V/18mA= 41.7 ohms. You need LESS resistance to reduce the idle current in the output transistors so try 33 ohms then 22 ohms.
This is with NO signal. With a signal then it will probably cause "crossover distortion" that produces a buzz on low level sounds.
When one bias diode is replaced with a wire then the idle current in the output transistors should be Zero and causes severe crossover distortion.