I think the conditions in which the car operates in will have a major effect (think wet and salt).
As I see this;
Wet conditions will tend reduce the heat produced by the brake simply because speeds will tend to be somewhat lower and braking distances longer. A wet disc will momentarily be lubricated with water, then very quickly turn that water to steam which will then behave similarly to the gas given off by overheated pads. But these affects will be very short lived and probably near non existstant with already warmed up brakes. Note that rotational speed will tend to throw water off the disc.
I can't think what affect salt might have...
I agree with John but I don't think discs or pads have to be hot. Some of the fade he describes is what is termed 'speed sensitivity'. Even discs and pads at ambient temperature will gas at speed when hard braking is called for. If a car does not have a sports brake set up, it is usually impossible to lock the wheels at high speed (say 80mph or above) yet the same pedal pressure can easily lock the wheels at speeds below 50mph. The point is pads can gas even when quite cool and this speed sensitivity is something that machined discs and more sporting pad compounds attempt to reduce.
'Outgassing' and the consequent 'pad fade' is a product of the temperature of the pad material, not of anything else. The pads will 'outgas' (smoke) only when they get too hot, that temperature being dependant on the pad composition. Pads get very hot very quickly when braking heavily from very high speed, so 'soft' pads can go from cold to overheated in just a few seconds if the braking is hard enough and the speed high enough.
'Soft' pads will 'outgas' at lower temperature than 'hard' pads, with 'soft' and 'hard' being references to the temperature at which the pads start excessively 'outgassing' rather than the actual hardness of the pad material (though there may be some general association between pad material hardness and outgassing temperature). 'Soft' pads (typical road pads) generally have a better coefficient of friction at lower temperatures, but 'outgas' more at relatively low temperature. 'Hard' pads (typical racing and 'sports' pads) generally have a relatively poor coefficient of friction at low temperature, increasing as the temperature rises, and outgas excessively only at a relatively higher temperature.
The "speed sensitivty" that Fruity is talking about (I think) is something quite different to 'pad fade'. As rotational speed rises the wheel gains an increasingly large amount of embodied kinetic energy, which becomes quite considerable at higher road speeds. The brake not only has to absorb the kinetic energy of the whole vehicle (from high to low speed), it also has to slow the rotational speed of all the rotating masses, and absorb that kinetic energy too. The rotating mass is substantial; the drive shaft, the hub, the brake rotor, the wheel, and the tyre (in the order of mass distribution away from the rotational axis, with the tyre being the most influential due to its' substantial mass and distance from rotational centre), so it takes a very significant %age of the brakes' performance to slow it's rotational speed.
At high road speeds a lot of the brakes' performance is taken up with absorbing the kinetic energy of the rotating masses, more so the faster the wheel etc is rotating. This leaves less of the brakes' performance available for slowing the vehicle mass, and so less available for 'locking up' the brake (i.e. overwhelming tyre grip). This is much more obvious at high speed than at low speed, because the rotational energy is far greater at higher speeds than at lower speeds. I think it is correct that the kinetic energy embodied in a rotating mass increases with the cube of the rotational speed (though it might be with the square of rotational speed).
This is why the pads get so very hot so very quickly when braking hard from very high speeds, i.e. it's not just the energy being absorbed (converted to heat) by the brake to reduce the vehicle road speed, it's also the absorbtion of the quite substantial rotational energy of the rotating masses. This is (IMO, and ABS notwithstanding) also why it is so much harder to lock up the brakes at higher speed relative to at lower speed, i.e. at high speeds so much of the brakes' performance is already being used to slow the rotating masses and so is just not available to overwhelm tyre grip. At low speeds much more of the brake performance is available to overwhelm tyre grip, so locking up is much easier to 'achieve'...
Keep in mind that after braking hard from high speed the pads are now very hot, and may well be 'outgassing' like crazy, so even low speed brake performance might now be seriously compromised (until the pads cool off).
This is another reason why reducing wheel, tyre, and brake mass is a 'good thing' in and of itself (i.e. not considering the compromises this might create). A reduction in rotational mass (and therefore some of the unsprung mass) not only improves handling and acceleration, but will also assist braking peformance, especially from higher speed (significantly more so than the same amount of mass reduction that is not rotational mass). On the other hand, increasing disc diameter will increase braking effect, so bigger and probably heavier brakes will have more braking affect at higher speeds, so the speed at which it is possible for the brakes to overwhem tyre grip should increase with larger brakes (especially if the disc mass isn't increased excessively as diameter is increased...).
In short, I think machined discs could improve high speed braking even without using a sport pad (which tends to have trade-offs in cold response, noise and disc life).
Grooved or drilled discs will improve high speed braking performance, if
the pads get hot enough to 'outgas', but not otherwise (i.e. if the brakes never suffer from 'pad fade' then disc grooves or holes are not going to make them better brakes). Such discs are more likely to be useful if the pads are too soft (as you infer), or with brakes that are subject to extreme usage and not even hard pads are fade resistant enough. There may be an advantage in fitting grooved / drilled discs insofar as their use may mean it is possible to use softer pads than may otherwise have been required, or lighter brakes than may have otherwise been needed.
I can't see much point in 'dimpled' discs, other than a very small increase in the disc surface area. In theory this will enhance the rate at which the disc can cool off (and disc temperature has a large affect on pad temperature as the pads 'lose' a lot of heat into the disc), though without data I tend to think the effect is likely to be very very small. I suspect that dimples are useful mostly because they look a bit bling, which helps to sell product...