Actually some Pedals will work just fine with 14v as opposed to 9v but not all...it all depends on the curcuit, If the curcuit is an all Opamp curcuit with no transistors or logic chips then many can even handle 30v or more depending on the opamp used, but with pedlas that have transistors in them the Increased Voltage will throw the Bias out and some logic chip will just fry with higher Voltages...I actually run my Opamp based pedals at higher Voltages as it increases the headroom but I wouldn"t suggest it to anyone unless they know what is actually going on inside the pedal....
True. Some devices have wide tolerances. Microprocessors, for example, have a maximum voltage, but can run at lower voltages if you also reduce the clock speed (or automatically reduce the clock speed when the voltage is lower, not sure whether the clock on modern CPUs is on-die these days).
LEDs require voltage limiting (well, technically it's really current limiting, but we talk about the devices as having a specific operating voltage usually in the 1-3V range and the current limiting effectively causes a voltage drop, so ...whatever...). Run them at a higher voltage/current and they can get too hot and the device eats itself or even explodes. Of course, this means you can run them at a higher voltage as long as you use a shorter duty cycle (a blinking bulb can run at a higher voltage if it is only on half the time than it could safely run if it were on all the time).
Transistors and other logic parts have a maximum voltage, but it is usually way above the circuit voltage. As you noted, though, it can change the sound.
AFAIK, resistors care only about the power drain across them, not the voltage, though I'd imagine there's a voltage at which the plastic becomes conductive, at which point you'd have a problem.
That said, there's a giant "but" there. Power is, in turn, dependent upon voltage. If the device on the other side of that resistor drains a constant amount of power (e.g. a resistor that provides power to a switching voltage regulator), you're fine. If the devices on the other side of that resistor drains a constant amount of current, then as you increase the voltage, the power across the resistor goes up and you could exceed the wattage of the resistor. This, of course, only applies to resistors in the power supply circuit, which are, quite frankly, fairly rare in modern power supply circuits except in high voltage electronics; for low-voltage applications, it is usually cheaper and easier to just use a voltage regulator.
Voltage regulators have a maximum voltage, but it is usually high enough (maybe 30V higher than the output voltage) that you won't run into it unless you do something really, really dumb (like hooking the 120VAC mains to a full wave bridge without a transformer and feeding 148VDC into it).
Diodes have a breakdown voltage above which they behave very differently when reverse biased.
If you raise the voltage high enough, you might encounter this, though with the exception of zener diodes, I doubt you'll reach it in most circuits. Most of the diodes I've seen lately have breakdown voltages of 35V or more.
Capacitors have a voltage limit. Worse, electrolytic capacitors lose their voltage tolerance over time. Thus, if you take a capacitor rated at 35V and run it at 12V long enough, its tolerance will degrade to 12V. If you then introduce 35V, the capacitor will explode and spit electrolyte out everywhere. This makes a horrible mess.
Worse still, electrolytic capacitors' voltage specification is based on a narrow operating temperature range. Exceed that temperature and the lifespan at or near the rated voltage is dramatically shortened. Since running a device at a higher operating temperature usually means that other components will dissipate more power as heat, this means you're doubly screwed.
It's not enough to look at a couple of components. You really have to look at the whole circuit and make sure that nothing is going to fry. Either that or talk to somebody who has done it before and find out how many years the device lasted before the caps leaked electrolyte.
