Well, maybe. But think about how long it takes a coin to clunk-clunk-kachunk through one of those machines and register its value. You might well be able to rig something up to auto-feed such a mechanism, but it would be noisy and slow, I think -- seconds per coin, rather than coins per second. I also seem to remember those mechanisms jamming fairly frequently, and I'd expect that to be more of a problem the faster you fed them. But, hey, if you can get your hands on one of the mechanisms, absolutely experiment with it!
What you suggest could separate silver from clad. The guts to a CoinStar machine would work. I believe they sort first by size, then by weight, and then maybe magnetism. Of course, this wouldn't solve the OP's quest to sort by date.
As luck would have it, I actually have a soda machine that is more trouble than it's worth. The coin acceptor works, but everything else seems to be dying one piece at a time. I've been debating on turning it into a gun safe for a little while now.. And yes, those are two problems to overcome - coins jamming and going for coins per second, instead of seconds per coin.
I wish I was the guy who thought of it first. But I'm terrible with coming up with original ideas. I've seen them for sale here and there. It seems like a good idea. Most gun safes are scary easy to break into. Older soda machines are a pain in the butt. You can get in them, don't get me wrong, but it takes time and a lot of noise. Unless someone knows something I don't. And just like with a safe, you can lag it to the floor. Every thief that's broken into my place has been way too lazy to put in the effort necessary to mess with it.
Just want to keep this thread alive, because the OP's idea is the GOAT. I'm ever on the lookout to automate anything and everything. Here are a few interesting links that someone clever might be able to utilize: https://coinoscope.com/ https://www.numiis.com/coin-checker/ It's possible that someone might be able to build a mechanical front end for either of these services. P.S. Neither of these services are very good. But, they're a start.
You can check patent applications, but if it was thought it would work, some coin grading company would have already bought the patent and would sue anyone trying to do it. Here is a source for some ( these are held by big time lawyers ) for the owner often a big money coin grading service we know by initials. https://patents.justia.com/patents-by-us-classification/73/163 Jim
It's worth checking back from time to time, because technology does march ever onward. In May 2022, the last time this thread was active, there weren't any AI systems that could do a credible job of carrying on a conversation, and look where we are now. "Recognize each coin's type, date, variety, and condition as it spills past a high-speed camera" might well just fall out from other developments.
Holy crap that's a lot of patents just to do with coin sorting! You're right, anything that could speed up the process of fishing would undoubtedly end up in the hands of a grading company. They'd stand to gain the most from it. I hope such a thing falls out. If and when I ever get around do it. I've got an idea for a ramp that will eat coins at a high rate of speed and poop out the silver ones into a separate bucket. Though, I plan to release the design with something similar to the GNU GPL. I invented something cool decades ago and thought to get rich off of it. It was a nightmare. Screw all that.
For more than 70 years, starting as a BM shop/show 9 YO assistant, I've dealt with coin collectors/sellers. Regardless of their worth, they generally are the most frugal? people I've met, willing to do virtually anything to acquire a believed free? cent. I'm a highest SAT test scorer Engineering-Specialist whose final job of 33 years was with a 90000 personnel international firm, solving first-generation design/build/install custom equipment world-wide problems. If you can do what you've proposed, I believe there are better uses for your talents/finances than try delivering an expensive device to a cheap audience who would want appreciable operational proof. You don't need to believe me, but I've dealt with some of the Creme De La Creme. JMHO
Good point -- but sometimes you end up doing something just for the challenge, even if it doesn't maximize profit/marketability.
Having multiple patents/records-of-invention/support-sources, I've quite often had to "prototype" principal portions of devices, or even the entire machine because of others inability to comprehend. A common question is "where did you learn that"? Often told by teams of "experts" that something can't be done as proposed! I spent >hundred grand to build a "first generation" proprietary drive-in sterilizer that more than halved the cost/time to bulk process, by using advancing MS/Phd engineering students (I wrote/taught an advanced college course for automated production) only, to help me build, by using available technology components. Once the available components arrived, the assembly took less than 2 weeks in "off-hours" to start operation, as others scaled/questioned my 1 assembly-drawing/bill-of-materials. The facility still does not understand, but accepted outcome. How does one explain to others, technology that has no proven basis? Don't, just use creative resources, as searching your ID (see: https://www.verywellmind.com/what-is-the-id-2795275) during rem-sleep, and awakening through a complicated training process. Immediately write answers found, that can't be accessed while awake, but the id allows retention for a short period! JMHO
This machine at least claims to be able to sort true coppers from zinc pennies: https://www.pennysorter.com/