Nope. I believe i accidentally solved the first. Tried a few times but get running into 51. Most i had was 56 but the endscreen kinda prevents you from analysing the board. The challenges is a great idea, like chess exercise in a complicated state, but desperately needs explanation and maybe a "the intended solution is 5 moves away" kind of deal. One can argue about how much control i have over the board, but for me I just send it without much strategy, since complicated board states like 1-2-1-2-1 have the obvious middle 1, but it all will depend on the rest of the board, launching flank attacks or maybe collecting om the side to protect corners. Its fun. Its complicated. But its a bit too complicated for me.
The initial threat issue... i believe there's no "fits all" solution. I think theres a video about "why hexagonal tik-tac-toe won't work" that explores the initial threat problem, but there are no solutions there i believe. I think you don't need to solve the flaw of the mechanic, but you might want to solve the player going in with a complete solution.
1) blocked squares - the initial blocked centre is good, but il lasts for 1 turn, not enough to prevent the player 1 from creating the first threat. Maybe multiple starting block patterns for 5 turns that prevent from using one solution. For example a checkerboard for 5 turns pattern will prevent player 1 from making a 2-1 construct on his third turn. Needs to be explored since player 2 will be able to put a threat. Maybe random cells unblocked gradually after turn 5.
2) player 2 getting a +2 on his first turn will prevent a one square
3) interest points (flat x2 cell or maybe every 5 turns it gives something. Or a cell that can be accesed (stabilised?) for a limited time and then it becomes innaccesable, or even becomes a x0 cell, voiding any operations done to it. Same can be done to some blocks on the edge of the screen, allowing risky plays.
4) random starting +- peppered in. Iffy about it as its just sort of a challenge map with a solution, but could be explored.
5) as i mentioned, abilities. The simplest i can think of is once a game being able to spend your turn to flip the value on one of the blocks, so if enemy accumulated a large number or accidentally cascaded into a large number, you could take advantage of that once, flipping a -9 into a +9. Maybe allowed on blocks >5. That would create a pressure to control the numbers and avoid accidentally creating cascades or avoid preparing big bully number attacks. Surely that can be explored to create baits or a person with better game knowledge setting up cascade laucher to make a big number and then crush the enemy's ability right away with another cascade.
Some things require new mechanics, some just extend the initial central square blocked mechanic. Up to you with what dirrection you wish to take it.
I personally have a strong feeling against "solved games" because my favourite childhood game was Mancala and this brings me back to it. If only it wasn't solved one could still enjoy it. If you haven't played it, there are a few online, you might want to try it, it has a lot of similar mechanics with cascading and setting up the board for the enemy or baiting them.
One of the issues of a solved game is the "complete information". I think it would be a different game mode, but a fog of war would be fun to explore, both players starring in one of the corners and blocked from seeing the other half of the screen, not knowing what the other player is building (or maybe where the other player's walls are and how they utilise them in their offence/defence)
Oooooh! I'll try the larger maps! WHY did you hide them from me!!! XD
Anyway. I hate puzzles, i like stategy games. I want to play it as a strategy game against complex bots. This game gets me thinking and i like it.
Good job.