I would like to get an expert opinion on the following issue:
Lets say I have some YUV420 8bpp footage (i.e. from DSLR) and the DP (me) set wrong white balance.
I want to fix it and of course the right way is to transfer the footage to xyY color space and alter the x and y (or even to have a prepared Kelvin offset slider) over the known color temperature curve http://upload.wikimedia.org/wikipedia/commons/thumb/b/ba/PlanckianLocus.png...
This process should be almost lossless.
Using lift-gamma-gain will give inaccurate (in terms of color) results and the footage will loose details (and noise will be added).
Is there a way in Resolve to correct color temperature properly?
Not by executing multiple colorspace conversions, for starters. At the end of the day, changing the inherent image values of source media is the bulk application of day-to-day color correction with software like Resolve. No matter whether you go to the extent of placing your original values within x'y'z' space or not will make no difference in the end result, as the same background operations need to take place -- at its most basic, gain adjustments in red, green, and blue, whether you are dragging a slider on a "color temperature" slider or rolling the gain and gamma balls. From a source-code standpoint, its just another algorithm like 'auto-balance' and either finds a home with your grade approach, or not, depending on your intrinsic control-comfort zone, ie "surrendering to the machine".
Leaving your source media in Y'CbCr in 709 space will not adversely affect the final result, depending on the original quality, Boltzmann electronic noise, and so on. If you have to boost blue, its going to get busy, that's physics. In any event, you are going to have to render the corrected media back to a colorspace and codec that everyday NLEs can deal with, so keeping it in its native locus seems like it should be the path of least resistance.
"I always pass on free advice -- its never of any use to me" Oscar Wilde.