I have been digging around a bit today and fixing them by yourself is going to be a hassle,
the resistors have to be huge and expensive to make sure they can take the heat, because they wil have to dissipate 40W, 80W and a whopping 120W heat to turn the motor at the correct RPM :-\
you could settle for something smaller, because they're actively cooled, but still, HUGE
I'm going to order a new resistor pack from the dealer, only 33 euro's
I made the mistake you are stating as well. It really is low watts. I have a 56ohm for the smallest one and it is only a 3 watts resistor- works better than oem by staying cooler, getting similar job done. I am going to go 3 watts for all of them if I have to. I am fairly certain ordering them can get it exact or very close to it. The big watts defy the resistance value, I am assuming in the resistor world, small watts to go with the algorithm is the output- not how much the other end can take. This still a guess on my part- I have narrowed it down to "whatever works."
I have to order some to complete this, will post results. If less than 1 ohm on a 200 ohm scale the ".5" for example is "of" 200, that is why I multiplied it, and the numbers really made sense after that. Less than 1 ohm resistor is not doing hardly anythjing at all- an ohm meter itself can have up to a .3 in just error, that is how sensitive that number is below 1.
It is almost an entire guess if it were'nt for the 56 ohm I just happened to have sitting around in the resistor block and functioning.