I need a way to saturate (not desaturate) colors in a GLSL shader. There's code all over the place for desaturating an image. Example:

vec3 desaturate(vec3 color, float amount)
{
    vec3 gray = vec3(dot(vec3(0.2126,0.7152,0.0722), color));
    return vec3(mix(color, gray, amount));
}

Many suggest converting RGB to HSV space before increasing saturation. However, I don't need to change hue, only saturation. If I pass negative values to the above function, it indeed appears to saturate the image. Is there anything technically wrong about doing it this way? Am I trying to take a dangerous shortcut here?

That code first computes a "grayscale color" from the input color (using the formula for Luma used in hdtv it seems) and then linearly interpolate between the input color and the grayscale color. Using that formula with negative amount you may get some component of the color greater than one. Depending on your needs you may want to clamp the result.