If you already know how 3d graphics works in general, and is experienced in reading documentation, I'd say you can pick up GLES in a few hours.
GLES 2.0 is really quite streamlined.
But you might be disappointed how low level it is, and how much code you need on top of it before you can do much interesting things
The difficult thing isn't really the API, but the 3d theory, lightning models, linear algebra, the computer science needed to write good data structures feeding the API, and things like that.
But wouldn't it be more effective to just try do it and see how far you get?
It's just a day, are you afraid you are going to waste it or something?
Never a waste to learn about new things you are interested in.