Been doing a little research and learned some pretty cool things. First of all when I examined the channels separately a found the grid was larger and more spread out in the red channel while it was small and tight in the blue. I thought to myself if that corresponds to the wavelength of light, maybe I can use it to calculate the thickness of the cmos cover glass?
Turns out the person who runs
Geoastro has done most of the work for me. Unfortunately they did not provide their spreadsheet so I had make a copy for myself to make the calculations a little easier. I will provide a
link for anyone who would like to play around with it themselves. Using it I calculated an average distance of 1.3mm, which sounds about right to me. It also appears I was correct in my assumption that an uneven overlap results in a round artifact in the final image.