That's interesting, and something I hadn't considered. Do you think the flow is less turbulent over long lengths, decreasing the psi/ft? And do you think it would be affected by whether the line is straight or coiled? (my kegerator isn't even 7 feet long).

I think the major factor is actually the turbulence induced by the keg coupler and the faucet/tap. Within the tubing itself, the flow should always be turbulent, and pressure drop should be constant. For a flow rate of 40 mL/s (a 12 s pour), Re ~ 10,000. Solving for pressure drop using a Moody chart gives f ~ 3.1e-2, so dP/l = 1.6e4 Pa/m, or 0.71 psi/ft.

So if you assume that a 5 ft line is balanced at 10 psi, the flow resistance due to the tubing is actually only ~3.5 psi, with the other 6.5 psi being dropped by the fittings. Ignoring the fittings gives a resistance of 2 psi/ft.

Try to carbonate to 16 psi assuming that same value, though, and you have problems because the actual pressure drop is 6.5 + 8*0.7 = 12 psi. There's 4 psi of excess pressure coming out at the tap. In order to balance that system, you'd need (16-6.5)/0.7 = 13.6 ft of tubing. That actually squares up pretty well with what I've seen in practice, so I'd guess that using the 6.5 psi drop for the hardware will get you pretty close.

Coiling the tubing shouldn't have any effect as long as the radius of the bending is large relative to the diameter of the tubing. I think that would always be the case for vinyl tubing, which you can't coil tighter than what, 6 inches or so?