I've relayed this story before, but as it's probably relevant.
For a while I was travelling around 40 miles each way for work. The start of the journey from home was ~3 miles of fairly clear slow driving, followed by ~18 miles of free flowing motorway, before hitting ~ 3 miles of crawling motorway (road works).
At the point I would hit the crawling motorway bit, my OBC would report ~45mpg. During the crawling motorway traffic, trundling along at engine idle speed in 2nd, the 'instant' mpg figure would be up in the 70s. During this 3 mile stretch the journey average MPG would get up to about 60mpg.
That's not mathematically possible since my first ~ 21 miles would have used about 0.47 gallons of fuel (@45mpg), so the next 3 miles would have needed to use a negative amount of fuel to hit 60mpg (0.4 gallons for the 24 mile mark).
The trip computer was actually averaging out it's journey MPG figure over time. Which is obviously nonsense.