To some extent your idea was attempted more than 20 years ago.

The first coin authenticator for the ANA was Charles Hoskins. When he left his position with the ANA he moved back to the Washington, D.C. area which is where I met him. He started the INS (International Numismatic Society) which provided a coin authentication service. Over time the INS started a grading service which I firmly believe was the first third party grading service. They started grading on an informal basis and then started issuing photo certificates as the demand for their grading service grow, but the first step was always authentication. If the coin was not real it was returned. Near the end of the INS? existence the ANACS and other services were providing third party grading and the problem with resubmitting coins for higher grades was in full swing. At this point in time Charles was approached by an investor who wanted to use a laser scanner to grade coins. The objective was to provide totally objective grading and to do away with the issue of knocking out coins and resubmitting them. A lot of money was invested in the project and they were able to demonstrate that once a coin was scanned it would always be recognized by the system unless marks were added to the surface (which would only lower the grade). The problem was in defining for the computer what exactly determined each grade. A speck of a given size on a three cent piece would be more of a defect on that coin than that same speck on a silver dollar. So each grade in the grading spectrum had to be tailored to the coin being graded. This effort provided proof of principle in that the computer could pick up all the imperfections on the surface of coins and recognize specific coins once they were scanned, but that is as far as it went. Personally I think the grading services should laser scan all submittals just to stop the resubmittal issue. The grading services should share this data bank.