Work on this is progressing very well. In a way this shouldnt be a surprise, after all, each board is a 2x5 array of 8x8 LED bicolour arrays, so it could be expected that much of the circuitry would be duplicated.
The first breakthrough was discovering that although each 8x8 array has under it a 14094 8-stage shift register and a 62083 8-ch Darlington sink driver, these are not a single elemental unit! Rather, the 8x8 LED arrays work as parallel pairs, top and bottom, and their associated chips are likewise divided into RED Cathode select, and GREEN Cathode select. The pairs of ICs under the top row of LED arrays handle the green cathodes, and the IC pairs under the bottom row of arrays handle the red cathodes - for both arrays in the column.
The result is 5 columns of paired LED arrays, with the green cathode selectors under the top array, and the red under the bottom.
So far so good. Next up was identifying that the two 14049 hex buffers are effectively the 'ends' of the data chain. One is where the data comes in, and the other where it goes out to the next board. Two of the lines from these are the strobe and output enable lines for the above chips.Two others are the data lines for the above - more on these later!
The other two lines from the hex buffers I have yet to positively identify, but one I think is the clock line.
The remaining ICs on the board, three 4015 dual 4-stage shift registers and three 62003 7-ch Darlington sink drivers, along with ten MJE210 medium power PNP's, handle the selection of the rows of anodes. All the 8 rows of anodes on each 8x8 matrix are in parallel with the rest of the arrays on that row. There are other connections involved here that I have yet to map.
Back to the data lines mentioned above. These were the big moment of understanding for me - one data line goes to the first shift register in the top row - with all the top row registers then 'daisy chained', until the data line reaches the 'output' hex buffer. The same occurs with the other data line for the bottom row. So in order to select a specific colour and column cathode, the data is serially shifted right along the board, and indeed the entire display, until the appropriate bits are all set.