While serving at a higher pressure does use more gas to push the beer, it won't make a huge difference in the number of kegs a 20# tank can carb and serve. If you compare 10-12psi serving pressure to say doubling it for a jockey box, you'd use ~22g more CO2 per keg. So if the 20# tank would have served and pushed 30 kegs at normal serving pressure, it would only carb serve ~28.5 kegs at double the pressure. IMO that's "nearly as many kegs".
And FWIW not all jockey boxes require higher pressure.
Not sure how you're calculating this, but I disagree again.
I do agree that the difference is 25 or so grams of CO2 if you go from 10 to 20 psi serving pressure, which is double the consumption if you don't include force carbing. Hopefully we both agree on that.
If you do include force carbing:
Let's assume you're carbing 19 liters of beer to 2.5 vol CO2 = 47.5 liters of CO2 at 20 degrees = 1.94 moles = 85 grams of CO2.
So if you force carb the keg then serve it at 10 psi, you've used 85 grams + 25 grams = 110 grams of CO2.
If you force carb the keg then serve it at 20 psi, you've used 85 grams + 50 grams = 135 grams of CO2.
If you have 20 pounds = 9,072 grams of CO2, that means you'd get 9,072 / 110 = 82.5 kegs if you serve at 10 psi, or 9,072 / 135 = 67.2 kegs. That's a 19% difference, which is very different from the 5% you calculated.
Edit: OK, I realized after the fact that you just took the 1-2 kegs per pound of CO2 rule and assumed an average of 1.5 to get your "30 kegs per 20 lb tank" number. I guess I'm just pointing out that if you want to assume an ideal situation (with no leaks, purging or other inefficiencies) you have to do it across the board. The 1-2 kegs per pound is a very empirical number, so you can't really modify that number with something calculated from the ideal gas law. The ideal gas law tells us that we should be getting 2-4 times more kegs per CO2 tank than we actually do, so obviously there are a lot of things we aren't taking into account. I'm just saying we should acknowledge that when calculating things like this. But my point still stands that in a theoretical (perfect) system, dispensing a keg at 20 psi vs. 10 psi will use significantly more CO2.