The whole idea of a ballistic finger print database is just a prelude to registering firing pins, bolts and barrels. Those are the obvious ones, but I'm sure that registering every individual gun part would be proposed. How much would a $7 firing pin cost once it had to have all the paperwork that a lower receiver gets? Anyway, since barrel wear is a factor, I'm sure it wouldn't be long before everyone had to be licensed to purchase ammo and submit information about the gun they intend to use it in. That way, if they get a 95% match to one of your guns , they could substantiate your claims that you've fired thousands and thousands of rounds since it was "fingerprinted".
If they had a database with hundreds of millions of guns, I think it would become much more likely to end up with a lot of false "certain" matches. Right now, if they have a recovered bullet or shell casing, a suspect, and a weapon, they can link the bullet or shell casing to the weapon and if they can link the weapon to the person, through paperwork or fingerprints, they have a case. However, with this "ballistic fingerprinting" it would be like going after someone because they have the same make of car and the same brand of tires that left tracks at the scene of the crime. When the sample size is small, the probability for error is small. As the sample size increases, the probability for error grows also. The database just makes it a lot easier to generate a suspect list without doing any actual investigation. Instead of being a suspect because someone saw you there at the time of the crime, you could become one just because someone in a lab said your gun's "fingerprint" was a 99% match with the one used in a crime.
They would likely come after you for making a tool "designed to help criminals commit crimes." Precedence doesn't matter in these cases. Look at the cities who tried suing gun manufacturers for making a legal product that works as intended.