Here's the data he runs, based on the "peak" horsepower available, which ignores that most vehicles don't frequently use their peak horsepower:

I decided to run the numbers for today's overpowered vehicle fleet. (The math is below.) Turns out we have something on the order of 51 billion peak horsepower sitting in our driveways. That's an incredible 38,276 gigawatts of power available. That absolutely dwarfs the nameplate capacity of our electrical power plants, which total up to a mere 1,087 gigawatts. In fact, each week of 2008, a horrible year for car sales, almost 38 gigawatts of capacity rolled into the streets of America.

Unfortunately, this is where things go awry. After pointing out that we were able to defeat Hitler with 80% less horsepower, which makes little sense, he comes to three conclusions.

The Tata Nano Argument
One, the current size and power of our cars and trucks is just stupid. The Tato Nano, with its 33 horsepower engine, is the way to go. (If all of the world's cars looked like that, going electric would also be a lot easier.) Let's merely note here that the average American passenger car has 7.5 times as much horsepower as the Nano and yet both vehicles will get you to the grocery store or to Nevada or wherever.

Technically, this is true, but with a top speed of 65 MPH it's not going to get there as quickly. Nor does the Tata Nano have to carry the same amount of safety equipment as a comparable car built in Europe or the U.S., so if that Nano gets into an accident the chance of injury is much, much higher (but none of his arguments work if you factor in progress). And if we're concerned about energy efficiency, the new Toyota Prius gets approximately the same mileage as the Tata Nano and will get there faster and with more comfort, more safety equipment, and more style.

The We Don't Need It Argument
Second, the people of just one hundred years ago would be awed by the amount of horsepower every American has access to. The funny thing - the irony, perhaps - is that we no longer need that amount of horsepower to do anything useful. The people of the prairie were scratching and clawing for every kilowatt hour of useful work they could wring out of some oil or the wind. The people of Omaha these days don't need anything like the direct energy services of their forerunners.

Yes, pity the poor people of Omaha with computers, air-conditioning, live-saving equipment like MRIs, and televisions. Clearly, it would be better if we went back to living without as much electricity.

The Insanity Argument
Third, and here's the hopeful part - no sane country would encourage its consumers to get on the technical and performance treadmill that led us to this point. Who would want this piece of the American technological infrastructure and set of consumer expectations? It's resource inefficient and expensive. I wouldn't expect the Chinese to follow our path to the American car anytime soon.

First of all, the Chinese are in fact doing all they can to mimic American cars and it is a growing market for American cars. So, they're apparently as crazy as we are.

But what's really missing here is that American cars are becoming more efficient in large part because of power increases. Automakers are using lightweight parts, direct injection, turbocharging, and other technology to increase power and, at the same time, lower fuel usage. The 2010 Ford Flex with Ecoboost (direct-injections plus twin-turbo) nets 355 HP, an increase in power of 93 HP over the naturally aspirated V6 model, while actually improving the mileage of the vehicle. We've engineered V8 power with V6 fuel efficiency.

Better efficiency is a good thing, but so is comfort, safety, and reliability. People in India aren't buying Nanos because they've made the conscious decision to limit themselves to 33 HP, they're doing so because they can't afford anything else. They're going in the opposite direction by trading the efficiency of bikes and motorcycles for something bigger. Americans can afford bigger, nicer, safer, more powerful cars for the simple reason that we, as an economy, have sought out progress, not rejected it. [Greentechhistory]

It's always been tough to stack up raw engine power output of the car against at-the-pavement performance. This has been the problem with dynos in the past, you either measure power on an engine dyno and don't know the power to the wheels or measure on a chassis dyno and don't know how much power the engine really makes. Rule of thumb driveline losses are never satisfying in either case and that's where Hyper Power International comes in.

Apparently, their dyno looks like any other four-wheel chassis dyno, with four big rollers to strap a car onto and sophisticated measurement equipment to test maximum output. The trick comes in when you've finished a pull and the system asks you to put the car in neutral. As the wheels and the rollers slow down, the dyno which was just used to measure power is now able to measure losses. By comparing the normal slowdown of the rollers against the slowdown with the car and all of its drag from gears, bearings, differentials and so on it can determine real driveline losses.

Motor Trend's Frank Markus was the first journo to test out the system and considering the constant automaker slap-fight surrounding the Nissan GT-R, Godzilla was the only choice for the test. The results are pretty interesting — it looks like the initial Nissan HP claims have been accurate. Of course, now there will just be naysayers wanting to see engine dyno versus chassis dyno versus Hyper Power dyno results now just to confirm Nissan isn't a den of cheaters. And by naysayers we mean Porsche. [MotorTrend]

Number 1: This is the electric motor. Running on AC power, it can rev up to 13,000 RPM while producing 248 HP and 200 lb-ft. Maximum torque is available all the way from 0-6,000 RPM

Number 2: The one-speed transmission. After reliability problems with a two-speed design, Tesla settled for this stronger, simpler one-speed. The 8.27:1 drive ratio allows for a 0-60 time of around 4 seconds and a top speed of 125 MPH.

Number 3: The Power Electronics Module. In addition to capturing the power from regenerative braking and using it to recharge the batteries, it also smooths out power delivery under hard acceleration.

Number 4: The battery pack. 6,831 Lithium-Ion cells are arranged in 11 series-connected modules and surrounded by sensors and cooling systems designed to prevent catastrophic cell failure. A full 53 kW-h charge takes about 3 1/2 hours, while all the batteries together weigh somewhere between 900 and 1,000 LB. Life expectancy is in excess of 100,000 miles.

Number 5: The body and frame. The Tesla roadster shares its extruded and bonded aluminum frame with the Lotus Elise while wearing its own, superlight carbon fiber bodywork. [via Treehugger]

The Best Value: The Shop-Vac - $21.60 per HP
At a price of just $108.000 (with free shipping), the 5-gallon Shop-Vac is a surprisingly great deal. You get a whopping five horsepower, which gives you a value of $21.60-per-horsepower. Of course, you can't actually drive it anywhere, but HammSammich was correct in pointing out the great value for a new product.

The Worst Value: AH-64 Apache Helicopter - $2,404.44 per HP
Powered by twin GE turboshaft engines good for 1,622 shaft-horsepower each, the AH-64A helicopter would seem like a great deal. Unfortunately, the delivery cost of $7.8 million cuts into the power advantage, and that's excluding development costs. The overall value is $2,404-per-horsepower, way too high for anyone but the defense department. On the other hand, it's the only thing on the list that comes fully loaded with an insane arsenal.

Best Land Vehicle: Suzuki Hayabusa Motorcycle - $61.85 per HP
There's no denying Picarso was right when he said that bikes are hard to beat in the bang-for-buck category. The 2008 Suzuki Hayabusa achieves an impressive 194 horsepower, all while costing a relatively inexpensive $11,999, for a total horsepower cost of $61.85 per HP.

Most Expensive Land Vehicle: Big Rig - $248.24 per HP
Though a used truck, like the one Unevolved found, offers a lot of horsepower for a low price, a newer truck is going to set you back. We spec'd a new Volvo 630 midroof truck and found that, with just 425 horses, the $105,500 price tag wasn't going to cut it. However, for that $248.24 per HP you get a vehicle that can tow anything else here.

An Actual Horse: $335.57 per HP
When Stoatmaster suggested an actual horse, we had to check and see how much power a horse truly puts out. A study in the July 1993 issue of Nature found that the max horsepower for a horse is actually around 15 horsepower. We looked at a lot of horses and found a pulling horse for $5,000, which seemed like a good deal. Unfortunately, at $335.57 per HP there's little value in a horse by this measure.

Best Car Value: Tata Nano & Ford F25 SuperDuty V8 - $75.76/$76.17 per HP
Though we thought it would be hard to top the Tata Nano, which only costs $2,500 after all, we have to give the Ford F250 SuperDuty and Eligh some credit. Though the Tata comes in at a low $75.76 per HP, the Ford and its low entry-level MSRP of $22,850 combined with 300 horseys is quite close at just $76.16 per HP. It won't look like quite a value when you factor in the price of gas, though.

Conclusion
While dollar-per-horsepower can be seen as a reasonable measure when comparing similar vehicles in the same class, there are too many factors that go into vehicles (and vacuum cleaners) to make it a true benchmark of value. That being said, this exercise clearly indicates the actual market value of horsepower compared to the other features.