Making my puzzle game Norna has been formative for me as a Designer, and looking back there are 6 design lessons I will take with me into future projects.
Don't force your game to be something it doesn't want it to be - let it grow of its own accord. When you have a game idea, it's natural for your imagination to go wild before having implemented anything. This is great for motivation, but it can also lead you down many dead ends.
Try not to introduce more than one new idea to the player at a time. You have to keep a check on the rate at which new ideas are introduced - you don't want to introduce a new idea until the player is used to the last one, but you also don't want to bore the player. This is especially difficult at the start of the game!
Players are constantly poking at a videogame to build their internal version of the game system. Having redundant mechanics or controls makes it harder for the player to form any clear idea of what the game is about. This is really about the pursuit of the most elegant design.
There has to be a clear understanding of what the graphics and sound communicate about the design. These communication "lines" shouldn't be obfuscated or stepped on when final art and sound are added to the game. You want the graphics and audio to be beautiful, but you also want them to be functional.
The player is giving you an account of the salient, subjective experiences during their playtest. It's important to not take this feedback literally, but instead to think about why they are saying what they are saying in the context of the entire experience (Hard!).
Just because a design decision was once right, doesn't mean it is still so after a year of development - after you've introduced a bunch of other ideas. Allow yourself to re-evaluate previously-good design decisions.