I get it now, but I don't think that will work the way I want it to. Example: 3 LEDs, wired in series for the white color. If I wanted LED #2 to change colors, turn off white and show green, wouldn't it need to be wired seperately, so there is control of the white? You wouldn't be able to turn off white and still provide power to LED #3, would you? If I picture the series, as like bridging with speakers, it wouldn't work. Hopefully the makes sense.
Yes, you are correct.
So if I am right, wouldn't I need to wire each LED, seperately,
Correct up to this point
...and regulate the voltage to 2V, for each, and require no resistor?
No. There's a part of this that has not sunk in yet. The LED forward voltage is approximate. It is approximate because, like everything, it has a tolerance, it is temperature sensitive, it changes with current, and it varies from one sample of the same part number LED to another. ALSO - any regulation you do is approximate - for the same reasons on the regulator. If you force a fixed voltage across the LED, the current can vary a lot from minor variations in the LED forward current and the regulated voltage. By putting some distance between the imprecise forward voltage and the imprecise regulator voltage and spanning that difference with a resistor, you greatly reduce the sensitivity of the current to the unavoidable variations in the two voltages. You can build active circuits, called constant-current regulators, to cut that sensitivity to almost zero, but you don't need to go to that extreme. A resistor is a good compromise.
The closer you get the forward LED voltage and the regulator voltage, the smaller the resistor value, and the more variation in the current/brightness/LED life you're going to get (all the way to the unwise and most sensitive situation where you have zero resistance and two varying, relatively imprecise voltages that come together at a single point - when they are forced to the same voltage, the difference between the ideal calculated and the actual current can get large). The trade off is efficiency where you can stack LED's to run multiple LED's off of the same power usage as running a single LED (original supply voltage x current in that string regardless of number of LED's in the string) - the higher you stack and the less the voltage difference, the more efficient you are, but the more current variation you will get with relatively minor voltage variations - hence the advice not to stack more than 4 or 5 volts from the supply voltage (you can get closer if your supply is well regulated).
You can see where putting the regulator in is a big improvement over running straight off of the car's hugely-varying system voltage. But if you're going to run only one LED in a single string, then you might decide that you don't need a regulator - that one higher-value resistor will be "good enough".
You've got to decide what you're going to trade off. You definitely need more than a "fixed" (a relative term - everything has a tolerance and variations over time) voltage across an LED - too many things you don't/can't control (like your car's system voltage and the LED's forward voltage) that are going to give large brightness variation and likely shorten life unless you cut your "nominal" operating current and brightness *way* back to allow for the variations. On the other hand, with one LED, it might be overkill to put in a regulator *and* a resistor - you likely can reasonably decide you can go with just a resistor (but definitely *not* a voltage regulator with no resistor).
To develop insight into those things, you can do the calculations based on certain assumptions for reasonable variations of the parameters. Reasonable conclusions on cars are likely to be that you need a regulator (plus resistor) if you're going to stack more than 2 LED's in a single string. For one or two LED's, you can get by with just a resistor - the most variation you will get is due to the difference between motor running (approx. 13.8 volts) and motor off (approx. 12 volts).