1 - What good would it do? Even if a drug dog sniffed drugs in a package, they know neither where it came from, nor where it is going. Perhaps AMS could have an internal policy whereby if a package is found to contain something illicit (even if it is the police who found it) the package is destroyed, so that the police wouldn't be able to track it to its ultimate location. I suppose the short answer is, the risk is there whether one uses AMS or not.
I'm more concerned as to how law enforcement would treat the owner of an AMS hub.
If I'm running a small business and moonlighting as an AMS hub, would I be held responsible for the contents of a package in my possession that has no identifying information other than an QR code with encrypted data? Packages at a FedEx Office location demonstrate at least an attempt by the shipper to identify sender and recipient to each package. Here's how I imagine it might go:
Agent: "This package is full of drugs! You're under arrest for possession with intent to distribute."
Hub operator: "Officer, that package doesn't belong to me. I was simply delivering it the same way the Post Office does."
Agent: "Okay, then tell me who sent it and where it's going."
Hub operator: "That information is in our computer system that suffered an unfortunate crash just a few minutes ago, erasing everything beyond retrieval."
Agent: "How convenient. Well, if you can't tell me who it belongs to, we'll have to assume it's yours. You're under arrest for possession with the intent to distribute."
2 - This one is more interesting. Remember the limited information on the QR code. The only thing the sending hub knows is whether or not their postage fee has been paid, the end hub city, and the recipient address. The only goal when onion routing the package is to merely conceal the fact that the receiving hub is the owner of the recipient address. You never know when you're at the last box because unless you can sign a message with your own address ... it may be the last box. Opening it would be just like opening someone else's mail. Why go through all that when you can just do your leg of the journey. The cheating hub wouldn't benefit anyway as all the postage is already paid out by the original sender. Best to just scan, throw in pile, and drive a couple hours to the next hub.
Ah, it's a decentralized system. I was picturing a single website to generate the payment addresses and create user accounts. So there wouldn't be an overreaching AMS arm to help identify hubs and recruit couriers while taking a cut of hubs' profits? I guess hubs could pay a "franchise fee" to get listed in an AMS directory and place Want Ads for drivers, etc.
3 - It would depend only on general volume. Remember, we're not only counting the packages going from San Antonio to Austin, but also from San Antonio to everywhere else that uses Austin as an intermediate hub between San Antonio and itself. It all starts be become very economical very quickly if there's enough volume to justify 2 or 3 times a day runs. How many of your Waco customers would pay $4 for same/next day shipping, if only they travel to AMS Waco to pick up their package?
Volume
is the key. Realistically, I only have about one customer a month within the state of Texas. Depending on how many bitcoiners were willing to try this, it could take a while to get off the ground. Still not a reason to forego it, just trying to be realistic.
4 - I could envision some system of record keeping, but I haven't really thought too much about that. It seems like a trivial add-on once the general project is up and running.
Seems to me potential customers might feel more secure knowing a courier can't run off with their package with no repercussions.
