I discovered this facebook group today. From what I can tell, it's real.
"******** Cab is a small-scale business of **** MA. The point of it all is to help out, typically the younger kids that are getting in some trouble. If you need a ride for a reason such as you are "under the [influence]" and don't want your parents finding out... a driver will come, pick you up and bring you safely home. The number is XXX-XXX-XXXX. The deal is to call that number and give your current address and the destination. There will be a charge of five dollars per person and you get wherever you need to go."
Although I don't know for sure, I suspect this may be run by an unemployed recent college graduate trying to make a little money in conditions of moral hazard - their parents will pay for major car repairs and all they have to do is pay for gas.
However, it does raise interesting questions about price-setting for things like taxis. This taxi service seems to pretty clearly be underpriced. Although you may take longer than usual to get home, you have a very price-insensitive customer base (setting aside the fact that the customers are drunk, they're trying to stay out of trouble and there aren't good alternatives), and a high fixed cost base (insurance, repairs) as well as variable costs (gas). They're helped by the fact that they have a very small geographic scale (a tiny town) so it's hard to fleece a driver for long distances, but it still is much less expensive than a cab would be, even if they're likely unlicensed.
Right now, taxi rates are regulated. This has both good and bad aspects - on the one hand, you know what you're going to get when you flag a taxi, and taxis can't (as easily) fleece unsuspecting tourists or drunk people.
On the other hand, actual price discovery is hard to do. Tyler Cowen has questioned whether taxis are allocated optimally, looking at critical mass of demand in various areas. Of course, this question becomes moot in a situation of dynamic pricing.
Certainly, a degree of standardization of taxi pricing is useful. For example, it would be very difficult to have one taxi charge by the mile and another by the half mile and another by gas consumption. However, if the rate structure was the same, but the per-mile component of the rate changed dynamically and was displayed clearly, you would fix the problems of "No taxis will go to dangerous areas because they're not compensated for the risk" and "taxis only go where people are accustomed to getting taxis, because it'll take less time to pick someone up". In other words, taxis currently minimize "chance of damage/crime" and something approximating "time between rides", which seems to result in a very different outcome than maximizing total utility. Dynamic pricing fixes this problem.
If taxis had to start at a certain base "first 1/8 mile" price (the initial amount - about $2 in Boston right now), but then were allowed to price each additional minute/eighth of a mile differentially (instead of the base "40 cents" it is now), then they'd be able to go anywhere and make it more worth their while. Pricing would not be perfectly efficient, because the base rate is standardized, but it's simple for riders and simple for taxicab drivers to manage. Areas with fewer cabs give you the prospect of better pricing, so you go there even if people don't usually look for cabs there. Competition is pretty fierce, from what I understand about taxis, so local monopolies would be hard to really create (especially given that a passenger could just get a cab to take them to a cheaper cab area with lots of supply before their longer trip).
This rate would have to be displayed prominently outside the cab, and perhaps shouldn't be changed at will (I pull over and immediately change the rate as you get in), but if a taxicab driver can say "For the next 15 minutes, I'm going to be in the East Village, where I can get this rate" and sets his "next 15 minutes" rate, and then switches his rate every 15 minutes depending where he is, then you have a situation where taxicabs could be allocated more optimally.