I've noticed this too, but I don't know if it's new in 2.2.8--I do remember seeing "NaN" in the color picker before, but that may have been when the device was new, before the attributes that comprise the color map (hue, saturation, and level) were populated. Regardless, it does seem like the picker could be pre-populated with the color if all parts of it are known, as they are in your screenshots.
As for the colormap implementation, the issue in that post seems that your device is reporting brightness/level separately from color, which is odd since an RGB value includes brightness as in inherent part of the set of values (e.g., 255, 255, 255 is white at 100%; 128, 128, 128 is about 50% gray, and...it gets more complicated for other values because of perceived brightness of R, G, and B, but similar idea ). I don't think this is a Hubitat issue--my biggest question is what the device means when it reports these separately. But one oddity on the Hubitat side is that "level" is called that when the color model--including the components of the color map--are the HSB/HSV model and not HSL, so just some non-conventional terminology (and range of values, though easily/linearly scalable to whatever conventional range if needed).