Myth Busted - Citadel colour mixing

For months now colour mixing charts are circulating on the web about a great chunk of the current Citadel range being possible to be replicated by simply mixing Skull White, Chaos Black, Blood Red, Enchanted Blue and Sunburst Yellow in the right proportions.
The story surfaced first at Nesbet Miniatures Blog and months later on the excellent Path of the Outcast Blog that I'm following. After sharing my concerns; Cannonfodder, the blog owner started a poll, asking the public what they think. 54% voted that they didn't try it but they thought it would work.
Actually it doesn't.
When I first read the original post at Nesbet's blog I shared the enthusiasm with all the rest, but after a little while things started to distub me a little.

A little on colours

For three years or so I was an active (and now honorary) member of Europa Barbarorum a project that some of you may know. It is a hardcore total realism mod for Rome (and later Medieval) Total War. Apart from skinning a few dozens of units in the mod, I was making some research into dyes and pigments, and how they work. (It was important to know what colours could be produced in the timeframe, and what was cheap or expensive.)
A colour has three major attributes: hue, saturation and lightness. It is possible to produce any colour by mixing rays of light in different colour, basically any hue can be recreated this way. What works with light doesn't work with pigments though. Pigments are of different opacity and will not produce all the hues you want. Of course I mean the pigments we modellers use; the cyan-magenta-yellow system for offset printing is an entirely different matter, and even there colours are not mixed, only dots (I know it is a bit more complex then this).
The other important thing is saturation. It is the measure how vivid the colour is. Pastel colours have more gray in them so they are of lower saturation.
The third thing to be taken into consideration is that pigments have personality... some hues you just cannot produce by mixing them. The best example is purple. In the antiquity a few grams of purple dye could feed a family for decades. Don't you think it was easier to mix or overdye madder and woad or indigo? It just did not work. Only purple dye did the job.
The charts don't, or very little address these points and this is the reason why they fail miserably.

The proof

No research is complete without experiment though, and it was relative easy - as all my readers know - I keep my citadel paints in dropper bottles. This time apart from cropping I did not use any Photoshop goodies and let the image much bigger then usual.
The results speak for themselves:

As ever, I don't pretend to have a monopoly for the truth, so if you thik that the colours are good enough proxies, or I did something wrong and you had different results, please do not hesitate to comment!
But from my part: Mith Busted.