[brucepick]

I also thought about a test.
I think it can be done using a ScanGauge.

(read to the end before you say "it won't work because...")

I think you probably have to drive some distance for it to give you any data you can use. If you don't have to drive any distance then the job is even easier than what's below.

With engine warm, reset the trip (segment?) and drive .1 mile.
Now can let it idle for two minutes.
Back in gear and drive .1 mile.
Get the mpg number for that trip.
Read the fuel consumed for the trip.
If it doesn't read out in quarts or liters you can derive it from the mpg because you know the distance.

Reset the trip segment for the next test.

Now repeat the process (drive .1 mile, stop for 2 minutes, then drive another .1 mi) but this time: During the two minutes you let it idle 10 sec, then off ten sec, then idle ten sec, etc.
Get the mpg for that segment and find the amount of fuel used.

Compare with the previous "trip".
Tell us how it turns out!
Probably should be done three times for each method for a more accurate result. Remember to start with a fully warm engine so warmup won't be part of the process.