This is what I wrote last week in a Beluga thread. I am not the final arbiter and freely admit I could be wrong. It's just my take on it. A re-post from that thread so I don't have to type it out again...
I call clean drives a trap. By all means use them if you like them, I did too for a long time. I focus on heat in my builds, and I believed clean drives were the right solution. I had the Orca, an AspX, DBX and even an Anaconda using clean drives! I know right?
One day I happened to notice that clean drives draw more power. And since power draw is what creates heat I wondered how much cooler clean drives actually are. So I ran some tests.
I had been under the mistaken impression that dirty drives would make the ship hotter all of the time, something I did not want in my cool ships. What I found was there is some increase in thermal load, but only when the thrusters are being used. Makes sense right? But what I failed to consider is that the thrusters are only used when speed is changed or the ship is maneuvered. So as soon as you reach a stable speed and attitude the thrusters stop producing heat and the ship returns to it's cool state.
Even then, the difference was a few percentage points of heat. Where it really shows is chain-boosting, where the additional heat stacks through the boosts. Once I realized this I re-did my ships with dirty drives. They become much faster, and since dirty drives draw less power than clean, some of the heat gain is traded off intrinsically. For a point or three of heat, the loss in speed potential is not worth the trade off in my view.
My clean drive days are in the past now. And as Ashnak noted, this does not apply to supercruise, so anyone using clean drives for say better scooping is not getting the benefit they think they are. It only has an effect in normal space, and even then, as my test showed, the difference is minor.