Can somebody explain why this preference?
Having had a GX8, which appears to process its images in the same way (though perhaps someone with both may correct me), it appears to apply an interesting process to the images.
What it will tend to do is reduce noise in channels which are locally weak and retain in in channels which are locally strong. So for instance, if you took a picture of a banana on a blue plate, you'd see a lot of blue noise in the plate (but little red or green) and a lot of red and green noise in the banana (but little blue). Personally, I found that very effective: it retains detail well and leaves noise which is mostly benign. In some cases it added character - shoot colour with a GRD/GX8 in a grey city on on overcast day and the colour 'texture' that's added by this algorithm can be appealing.
However, there's no denying that most people who shoot colour will object to - for instance - noise in skies. Most JPEG engines in cameras will smooth out all noise, which makes for smoother areas of colour, and the GRD/GX8 algorithm retained more noise than most people like.
This is especially as the ISO started to go up, and whilst the resulting qualities could be appealing at base ISO, the sheer amount of noise as ISO went up could be an issue, especially if you didn't want to use an aftermarket noise application. Dynamic range suffered as well - something that generally hurts colour images much more than B&W.
In B&W the noise pattern was just nice, though - it was quite nebulous, free of the typical digital edges and patterns that you often see at the pixel level. This meant that even when there was quite a lot of it, it wasn't ugly and it didn't rob detail.
--
http://flickr.com/photos/stewartpratt/ 
