I've had complaints of poor mileage after changing a couple people belts early on. I've always use the correct tools. That said, it is possible to have everything correct as far as the manual and how-to goes, and get reduced mileage.
You read on here over and over again that the timing is optimized at the top of the graph. This is true...for the most part. A few comments:
With the engine warmed up, correct engine selected, etc. and the timing set in the upper 25% of the graph, The timing will be off. What!??? Yeah really.
If you look at measuring block 4, actual timing vs requested, You'll see that the actual is somewhere in the 2-3 BTDC range, while requested is in the 0-1 BTDC range.
Everything is correct, but the IP is at the limit of how far it can retard. Yes, with the static timing all the way at the top, it can reach more advance, but realisically, you're going to be driving at light load on the freeway a lot more than you'll be at 4K+ RPM @ full load.
At idle, this has very little effect on the fuel economy. At light load on the highway - cruise on, 70 mph - the timing will be in the 1.5 BTDC range. If the car cannot retard below 3 BTDC, you'll knock a few MPGs off. The more time at light load, the more it will effect the economy.
So if you've dilligently followed the proceedure and you have the static timing at the top of the graph, - put it in the middle and see what you get.
Also, after changing many, many belts these cars, I think the best practice is to:
Warm the car up
Check the timing
Advance the IP as far possible without making the cold start/advance solenoid duty cycle (again, block 4) rest in it's full retarded position. If the cold start solenoid duty cycle value is 2.8%, it's completely retarded.
I shoot for the 15-25% duty cycle at warm idle.
This is usually slightly above the blue line in the timing graph; depending on ambient air temp, barometric pressure, and fuel quality.