I realize this issue (warping) has been repeatedly addressed on this site. I've just graduated to high-temp filaments (PC in particular). I don't know much of the physics of this. I'm wondering whether the degree to which the filament contracts is proportional to the amount that it cools. If the answer is yes, then wouldn't it suggest that a lower printing temperature might reduce warping-as the temperature interval over which the filament cools is smaller? Or perhaps the difference is negligible?
Also, I see a lot of emphasis placed on good first layer adhesion. Is this still an issue if you are printing on a raft?
I can't address polycarbonate specifically, but can provide a general overview of the higher temperature filament considerations.
Printing on a raft means that the adhesion temperature of the filament is accomplished. This temperature is the factor to be considered if you are thinking of dropping the printing temperature. If you drop below recommended minimums, you risk losing adhesion to the build plate and also inter-layer bonding. That alone means one should use caution when dropping printing temperatures.
Printing with a raft usually means the model's individual parts have such a small footprint that they would not remain bonded to the build plate. Rafts are also used on printers with an uncertain planar surface or irregularities in the surface. That's not applicable to this question, generally speaking.
Your question about contraction being proportional to the amount of cooling is perhaps misdirected. One could consider that the printing temperature is a manufacturer specified value and the cooled temperature would be generally considered room temperature. Room temperature would be addressed as a range, rather than a single value, but even as a range, there isn't going to be a big percentage of variation in the calculation involving the print temp/room temp.
My experience with the higher temperatures is more related to the volume of material per cross section (in all three dimensions). A printed model of substantial height with a relatively small horizontal cross section (think cylinder) is likely to have much less distortion in the x/y plane and greater distortion along the z-axis. The mass of filament cooling in the z-direction generates greater force than the smaller mass on the x/y axes.
Another factor in such thought processes is that layers are on the x/y axes and the strength of the extruded plastic is more homogeneous through the nozzle, while the z-direction creates inter-layer discontinuities, making warping and delamination easier.
I've found that I can reduce (but not eliminate) warping and delamination if I am able to maintain chamber temperature for longer periods and reduce temperature slowly. Unfortunately, I have a semi-enclosed printer and the heat loss is dependent partly on the ambient air temperature. A fully enclosed heated printer with auxiliary heating under some form of control may give you the best results.