No, it isn't the same effect as adding resistance, and you aren't feeding 12 volts down the 5 volt line!
Adding resistance only adds a voltage drop if you are drawing current - so you need to know the running resistance of the fan. The resistance will dissipate the lost energy as heat. A perfectly acceptable way to do it - but be careful where you place the resistor.
No look at using the split rail method.
Lets say the fan draws 120mA at 12 volts. We are looking to loose 5 volts at 100mA, so the fan's dynamic resistance is 100 ohms. We need another series resistance to reduce the current to (using ohm's law) 7/100 ohms or 70mA. The resistor will be 5/.07 or about 70 ohms, and will dissipate 25/70 = .35Watts. My figures are example, the figures for an individual fan will vary. Resistors come in standard values - in this case the nearest values would be 68 ohms or 75 ohms. http://www.logwell.com/tech/componen...or_values.html
The 5 and 12 volt rails are reference to a common earth or ground rail - so it is a potential difference that exists between them.. The only possibly down side is that it might marginally affect the regulation of the 5V rail.
Consider (as an example) a power supply with a common earth and a plus 12 V rail nand minus 12 V rail. The voltage between the two will be 24 volts - and you could power 24 V equipment from it without using the earth point at all. This is exactly the same case - except both rails are 'above' ground potential so the potential difference is only 7V - the difference bertween the two.
I cannot comment on the problem you experienced with the burnt out cable. The only occasion I have seen that happen is with a cheap PSU and a mis-connected floppy header - so perhaps in your case the pug had been displaced or there was a wiring fault in the adaptor. However without looking at the cct diagram of the components and their physical configuration, it would be impossible for me to make an accurate diagnosis.