I want to comment on "Update 1.1 : add a checkpoint and rename another to improve one of the challenge. 1.4 : Rework of almost each puzzle to be a bit more accessible and hint in the correct direction."
specially since it's the second game in a row I came into today with a Patch n. 1.1 made to address this problem.
Fact is, programmers and game makers are so smart that they have a hard time empathizing with the bulk of their player base -- even the player base that interests themselves in a game like this -- when they are faced with the puzzles in their games.
We see this all through the first decade, and maybe the second too, of PC games, when the PC world was insular, and (I am thinking of the 1980-1990 span of time, mostly) those who played the games were much alike those who made them. Think of adventure games dating to that period.
We stopped seeing it when games became an "industry": salespeople and marketing folks made sure to enlighten game designers about the abilities of most players. (Granted, the most talented players find such games almost offensive, but they have no weight in the sales charts, have they.)
Nowadays we meet with such a problem (too tough condundrums) only in Indie games, made by programmers who, for the most part, are used to talk and deal only with other programmers, and with, well, their programming tools.
Then poor sales and people's complaints come, and 1.1 hardship-watering patches along with those.
But I say: why not do what they did with The Last Express' Steam port? eave your creation as you felt it had to be originally, just add a 3-tier help system.
When the player is at the end of their wits and wants to chuck the towel in, he/she asks for the first, more indirect, hint. If they wont more help, they click again, and grade-2 hint comes. And a third click delivers the straight solution.
This is the system that, while it leaves authors free to be themselves 100%, matches the needs of as wide a player base as possible.
Why so few games have it, I wonder.