Pulling from that topic:
As a tester I didn’t want to spill the beans until post-comp.
I actually really enjoyed figuring out the rules. Maybe I’m more suited to this sort of puzzle, and I enjoy the meta-puzzles beyond “yay! I got it right!”
But I simply wrote down my first few guesses. Then I said, there’s probably logic here, but what is the logic? Of course, it’s a bit unrealistic to have it the way it is, but I enjoyed the twist.
Coming back to this topic I’m glad to see a mechanism planned to spill the beans on the rules, but only if the player wants it.
As a tester, I enjoyed trying to get things right in fewer than 6 guesses based on the abstract clues and on knowing I was looking for a real word. It added a wrinkle to the usual binary search, which had been done in many ways.
I also really liked a moment of realization I had for the guesswork that nnnnnn was a beter first guess than mmmmmm, as you were more likely to get a letter right.