Also, the color channels are very much correlated with other channels, including height, including the regular pattern bits. For instance, red and long or yellow and short coexist in 103 cases out of 152. Zbyszek2 sampled WIF keys and performed analysis and proved in his post earlier, that it is impossible to see such a correlation in real WIF key.
The 011 pattern could be very well
only left there for us to properly read the bits (cause you can only read it in 2 ways, following same order just in different directions). Next, if you do a data analysis against Heights, then I have some bad news for you - from the fact alone, that there is a constant pattern of 011 in every other bit, your analysis gets corrupted begin with. There will be 25% more long ones, and given the psuedo-randomity of other tracks, this can very well push the 76/152 to 103/152.