It is usually used to describe a riskless transaction.
In the academic sense, yes. In reality, no.
Correct. It's called, "market making".
It's not terribly scalable because you need to have transport and time to attend auctions, plus somewhere to store the stuff. But if you have a decent sized car/pickup, proximity to auctions and a nose for a bargain, you can pick up some good pocket money with this strategy.
Incidentally, I think the market for the pricenomics data is precisely the same as the KBB market - detailed lookup via mobile app for people who are buying/selling regularly. A good pawnbroker/2nd hand dealer/auctioneer will know the price point on a wide range of goods, but allowing people to purchase this information for a reasonable price will help both people starting out, and also professionals stepping outside their bread and butter.
I would suggest a freemium model, where the free version gives you a price range, and the pro version gives you a very direct price.
This is why you buy in the small and isolated market (craigslist) and then sell in the larger market (ebay) where you don't have to find people and waste time and money delivering the item to the person.
1) The Toy Story 3 plot: people want to find a home for their possessions with people of similar interests. They become angry that you are reducing the consumer surplus that would have gone to the ultimate buyer.
2) They realize that arbitrage on a first-come-first-served marketplace can reduce the value of the marketplace as a whole, since buyers are less likely to find bargains. They recognize your behavior as antisocial and become angry.
Friend of mine does the whole estate sale book thing, too. People have definitely gotten upset when they realize what the scanner is for.
I wrote a program awhile back that would look for iPod Touches on my local Craigslist and would analyze the price for many different versions. If an individual iPod was available and was below mean price within a given time frame, I'd send the seller a note almost immediately. Then, if I purchased the item, I'd relist it at the mean price. Completely an arb play.
What I learned was that 1) iPods are easy to sell, 2) People who buy on craigslist are flaky and tend not to show up, 3) The time spent waiting around to meet people was obnoxious, and 4) Sometimes sellers see what you did with their item and send you angry emails.
I never made more than a few hundred bucks, but it was an interesting experiment nonetheless.
Second story: For years, my wife sold used books. She'd go to estate auctions and buy boxes and boxes of books. She'd go to library sales and buy hundreds of books. Then she'd sell them on the Amazon marketplace. She would walk into these places with a barcode scanner and would check their price and sales rankings on Amazon. If it was in a decent margin for her, she'd buy the item. This actually did scale, but she quit when it became less of a hobby side income and more like real work. She's a nurse by trade, and the book selling was never meant to be anything more than a hobby.
So, I guess what I'm trying to say is that arbitrage of used goods can be a business, but it depends heavily on the availability of the supply of used goods. Books worked for my wife because she could buy dozens of them at a time and ship dozens of them them all from the post office down the street as orders came in. For me, I was running around town picking up items and dropping them off.