Not as good as it could have been
It's not that this thing isn't good at all. It's a perfectly normal platformer. The thing is that I think you've stumbled upon a great fantastic idea, but haven't quite done it right. The concept of the colour mixing is what could have really sold this game.
But it falls short somehow. I think the main problem here is that there is too much 'platforming' rather that 'puzzling'. And I feel it's such a waste because there could have been a lot more puzzling going on with the concept you came up with. There could've been much more problem solving involved in levels instead of merely avoiding stuff and timing right.
For instance, levels 28 and 29 were just totally unnecessary. They didn't even have anything to do with the idea of color swapping to solve the levels. In levels like those it was more of just an ordinary average jump and avoid platformer and the colour changing was just something that got done along the way.
You should concentrate more on the puzzle element. Design levels that require thinking and methodical application of the colour swapping. Levels like 26 and 27 were I think the best representations of what this game could have been. They were tricky and required some sense of thought before randomly taking certain colour changers. Especially when it came to the secondary colours.
A lot of really good puzzle games fall victim to this and I'm sure many gamers who are keen on puzzle games for the sake of the more intellectual level of playing always share the same frustration when there is a level that requires either impeccable timing, or skillful jumping - none of which are required in real puzzle games. The act of solving the puzzle is hard enough and enjoyable enough. There is no need to rely on regular platforming challenges.
Concentrate more on the puzzle element. The concept is a good one and really attracted my attention. But most of your levels don't use this to it's fullest potential and I think that is quite a let down.