At roughly 50 MPG on average it seems to make a lot of sense. But I have a pretty good car already, my Saab averages over 30 MPG in all of my driving which combines a lot highway and a fair amount of city. [This is a lot better than it is supposed to according to the EPA estimate, not sure why]
So here is the thing, assuming I drive 15,000 miles a year (this is high, but for argument's sake...) I will save about 200 gallons of gas. So, at $4 a gallon this saves me about $800 a year. Add another $1 for the cost of my carbon consumption (reasonable figure according to some estimates) and the total comes to $1000 a year. This is considerable, but not as good as you might think given the additional 20 miles on a gallon of gas I gain.
And thus I run into an error of logic that a lot of people have been talking about recently: going by the MPG numbers gives you a distorted sense of the benefits of fuel efficiency. To see this, consider if I had a car that got 15 MPG and switched to a very modestly better car that got 20MPG. Over the same 15,000 a year, I would save 250 gallons of gas!
The moral is that all the people that trade in their reasonably efficient car for a Prius, don't do nearly as much to reduce carbon emissions as a person who went from a Suburban to a regular sized SUV - or to a Prius. [Of course this doesn't factor in the secondary market in which poor fuel economy cars are becoming incredibly cheap - encouraging people to buy them]
It doesn't meant that I shouldn't buy a Prius and when I my current car wears out, I will look to find a fuel efficient alternative, but in a few years I hope to have even better options.