Question: As far as evolution is concerned what natural mechanism would account for keeping the "Scorer" Happy?
Actually I can think of one, but there must be more.
How about certain "tries" cause your construct to degrade or fall apart? There will be no "memory" of the error, however, unless Alamo-Girl can find one in there.
If that is the same program I've seen before, it is invalid as it correlates a pseudo-random selection to the known content of a position and then fixes it and moves onward to the next position. That is not random by any definition or stretch of the imagination.
A much better test of "happenstance" in evolution would be to digitize Hamlet and then kick off a random number (string of numbers) generator and tally the iterations required to come up with Hamlet by happenstance.
Of course you could reduce Hamlet to a number from its digitization series of numbers and then compute the number of processor iterations required to arrive at that number on the same processor used above. This would be a baseline.
Or you could lay down the pure happenstance insistence and take up the information theory approach to computing the evolution of biological information: Evolution of Biological Information