The accent helps, but not quite enough IMO. Maybe a bolder accent? When the image is shrunk down (for hats, polos, etc) the eye kind of gets lost and is hard to differentiate from the other "random" sections of white and green on the head. Perhaps a couple of those need to be removed? Especially the white ones near the eye. Maybe try removing the white spots above and below the eye, and the left half of the one to the right of the eye. I think when it's shrunk down, the eye needs to be clear to avoid the appearance of it being just a black blob with random spots of color.