A few months ago I published a swatch of various red/magenta colours. Bruce McEvoy of Handprint remains sceptical of the lightfastness claims for a number of reds and always suggest you do your own tests. Following his advice this is the result.
The swatches as originally painted
The same swatches after four months exposure in a south facing window. I put masking tape over the bottom half of the swatches so any deterioration should be in the top half. Personally I can detect either very little or no difference. You have to allow for the fact that reproduction may not be of the very best, but I have the original and looking at it reached a similar conclusion. I don't know what the optimum time scale is for doing such tests and I suppose they might well fade over a much longer period, but then again watercolours are not normally hung in south facing windows so make of it what you will.