Somewhere I read that the Model S calculated power efficiency gets somewhere around 75 miles per gallon. If this is so, how is such a calculation and comparison made?

As other poster mentioned, it's more like 90 MPGe, based on EPA data. What the EPA did is take the amount of energy a gallon of gasoline gives off when burned and convert it to electricity, which gave them 33.7 kWh per "gasoline gallon equivalent."So 90 MPGe means that the electric car will go 90 miles on 33.7 kWh from the wall (including charging losses) on the EPA test cycle.The price of electricity varies widely from location to location, but is generally fairly stable over time, and averages a little lower than gasoline on a per unit energy basis - so the metric that matters, miles per dollar, can be either better or worse than this number suggests depending on where you are. And of course, this is an EPA estimate, so as always, YMMV. Walter

I know this has been discussed before. I think perhaps the flaw in the EPA analysis is that it does not account for the fact that only about 30% of the energy produced by combusting gasoline gets to the drive wheels, whereas in a Model S over 70% of the energy drawn from the battery gets to the driven wheels. Perhaps someone more expert than I can confirm that.

For the Model S, it is a whole lot more than 70% that reaches the wheels. I'm not at all convinced this is a flaw of the analysis. It is certainly reflected in the analysis, but it is a valid part of measuring the car's efficiency, which is what EPA mileage estimates were created to do. A car that gets less energy to the wheels costs you more to run and wastes more energy, just like a heavy or unaerodynamic car does. The power company doesn't sell you power by GGe, which may make it a like hard to see the cost comparison (which varies quite a bit based on your location and power plan type anyway,) but the comparison is completely valid as far as I can see. Walter

This is a big problem with their assumptions. They don't take into account the fact that most gasoline engines are only 25-30% efficient whereas electric motors are 90+% efficient. They also don't take into account that it takes another 25% of oil energy to refine and deliver gasoline to you so the efficiency of gasoline is 25% worse. I take a different approach to calculating MPG equivalent. I assume that a large sedan like the Model S, if powered by gasoline, would get 20 MPG. At $4.00/gallon, that's $0.20 per mile. The Tesla uses about 350 wh/mile which costs me $0.03 so the same 20 miles would cost $0.60 per gallon... or alternatively, for $4.00, I could go 133 miles. (Your assumptions and costs may differ... but this works for me.)

Another way to figure it, after you have some miles on your Tesla, is: Odometer reading equals mpg divided by 1000. Then calculate the amount you've paid for electricity for that distance and compare it to what you would have paid for gas. Example: 25000 miles on Tesla. 25 mpg gas car would use 1000 gallons of gas. Then it's easy to do the math based on the price of gas. This method works best if you have never reset your trip counters because you can read the kWh used, multiply by your electric rate and add a bit of a fudge factor for charging inefficiency.

For the purposes of what the epa is stating, it doesn't matter of the tank to wheels efficiency is 1% or 99%. What they are telling people is how much fuel they need to put in the car to make it go X miles. If you want efficiency of the drivetrain, what you suggest would be great. But the EPA numbers are not about drivetrain efficiency.