Yep, estimating city v highway miles for each tank is just that, an estimation, i.e. a guess, but at least it gives you some inkling of the environment in which the miles are being driven.
In my case, I stipulate a standard 10% city for each tank unless my driving conditions are significantly different for any given trip, in which case I stipulate a different proportion. I generally use 10% city miles driven as my standard split because that's very roughly the proportion I drive in one small and two large cities, and the rest is on the highway. That's the best estimate I can give for each tank and still keep my sanity.
I've found the largest influence on my fuel economy is the speed I maintain on the highway, since 90% of my miles are spent there. So, my stipulating that I spend 90% of my miles on the highway gives a reason for my higher MPGs as compared with someone who spends 90% of their miles in the city.
If I were to stipulate average speed it would further refine reasons for the MPGs I achieve, but since my car does not give me that information, I'd have to keep notes on elapsed time during trips, which would be quite the headache for me.
So even though I think you're right that listing the average speed would be worthwhile in comparing fuel economy between cars, it would cause a lot of complexity for those of us with cars that don't provide that information, and especially for those who spend a lot of time in the city with multiple short trips for each tank.
__________________
__________________
|